Bhel - Transformers-Tata McGraw-Hill Education (2003)

TRANSFORMERS
Second Edition
C HAPTER 29
About BHEL
Bharat Heavy Electricals Limited (BHEL) is the largest engineering and manufac-
turing enterprise of its kind in India and is one of the leading international compa-
nies in the field of power equipment manufacture. The first plant of BHEL was set
up at Bhopal in 1956, and currently its range of services extend from project
feasibility studies to after-sales service, successfully meeting diverse needs
through turnkey capability. The company has 14 manufacturing units, 4 power
sector regional centres, 8 service centres and 18 regional offices besides project
sites spread all over India and abroad.
BHEL manufactures over 180 products under 30 major product groups and
caters to core sectors of the Indian economy—Power generation and transmission,
industry, transportation, telecommunication, renewable energy, etc. BHEL has
acquired certifications of both ISO 9000 and ISO 14000 standards for its operations
and has also adopted the concepts of Total Quality Management. BHEL has
adopted Occupational Health and Safety Standards as per OHSAS 18001. Two of
its divisions have acquired certification to OHSAS 18001 standard and other units,
too, are in the process of acquiring the same.
BHEL has
• Installed equipment for over 62,000 MW of power generation—for Utilities,
Captive and Industrial users.
• Supplied 2,00,000 MVA transformer capacity and sustained equipment operat-
ing in transmission and distribution network up to 400 kV—AC and DC.
• Successfully tested transformers of rating as high as 200 MVA, 1-ph generator
transformer and 167 MVA, 1-ph auto transformer at an independent lab in
Europe.
• Supplied over 25,000 motors with drive control system to power projects, petro-
chemicals, refineries, steel, aluminium, fertilizer, cement plants, etc.
• Supplied traction electrics and AC/DC locos to power over 12,000 km railway
network.
• Supplied over one million valves to power plants and other industries.
TRANSFORMERS
Second Edition
Bharat Heavy Electricals Limited

Bhopal (MP)
Tata McGraw-Hill Publishing Company Limited

NEW DELHI
McGraw-Hill Offices
New Delhi New York St Louis San Francisco Auckland Bogotá Caracas
Kuala Lumpur Lisbon London Madrid Mexico City Milan Montreal
San Juan Santiago Singapore Sydney Tokyo Toronto
Information contained in this work has been obtained
by Tata McGraw-Hill Publishing Company Limited,
from sources believed to be reliable. However,
neither Tata McGraw-Hill nor its authors guarantee
the accuracy or completeness of any information
published herein, and neither Tata McGraw-Hill nor
its authors shall be responsible for any errors,
omissions, or damages arising out of use of this
information. This work is published with the
understanding that Tata McGraw-Hill and its authors
are supplying information but are not attempting to
render engineering or other professional services. If
such services are required, the assistance of an
appropriate professional should be sought.
Tata McGraw-Hill
© 2003, 1987, Bharat Heavy Electricals Limited, Piplani, Bhopal, MP
No part of this publication can be reproduced in any form or by any means

without the prior written permission of the publishers
This edition can be exported from India only by the publishers,

Tata McGraw-Hill Publishing Company Limited
ISBN 0-07-048315-9
Published by Tata McGraw-Hill Publishing Company Limited,

7 West Patel Nagar, New Delhi 110 008, typeset in Century Schoolbook
at Tej Composers, WZ 391, Madipur, New Delhi 110 063, and printed at
Rashtriya Printers, M-135, Panchsheel Garden, Naveen Shahdara, Delhi 110 032
RZZCRRDDDZRRL
The McGraw-Hill Companies

To
The Transformer Designers
at
BHEL
C HAPTER 29
C HAPTER 29
Message I
It is indeed a pleasure to see the revised edition of this book on

transformers, covering various aspects of design, manufacture and
operation.
This book, brought out by one of the leading organizations in the
transformer manufacturing industry in India has been in print for
the last 15 years. Although originally intended as a reference for
practising engineers, this book is invariably used in under graduate
(pre-final and final years) and post graduate classes. The fact that
this book is so popular and caters to the demand in the transformer
industry is a tribute to the publishers, BHEL and the authors. The
inclusion of chapters relating to the latest developments in the field,
viz. condition monitoring, life assessment of the transformer, con-
trolled shunt reactors, short circuit design aspects, etc. along with
updation of the existing chapters keeping in view the latest trends in
design, material development and testing, and solved examples in
the revised edition shall aid in the better understanding of the sub-
ject by the readers, especially the students and upcoming engineers.
I find the work informative and exhaustive and recommend it as a
hands on practical guide.
I sincerely hope that this book serves as a useful reference not
only for the present generation of transformer engineers but also for
institutions and students.
Dr. Sankar Sen

Ex Power Minister
West Bengal
Kolkata
and
Formerly Vice Chancellor,
Jadavpur University
C HAPTER 29
Message II
I am pleased to know that the book on Power Transformers, cover-

ing various aspects of the equipment, written by experts at BHEL
has been revised and enlarged to include new apparatus like the
Controlled shunt reactor and practical aspects like Site monitoring.
The book in one volume treats all the major aspects of the trans-
former. It explains the operating principles, design, materials, mag-
netic, electric and thermal circuits, fittings and accessories, testing,
erection, commissioning and maintenance aspects of all types of
power transformers and reactors, including special transformers,
rectifier and traction transformers as also HVDC transformers. The
content of the book has been ably presented with an original and
vivid approach.
BHEL has rich experience in the design, manufacture and supply
of large power transformers. In fact the transformer is one of its star
products.
The book should be of great value to practising engineers, re-
searchers and students in the area. Apart from individuals, organiza-
tions in the power sector will also find it a great help in understand-
ing the intricacies of power transformer performance and testing,
and their impact on the economics of power delivery systems.
It is anticipated, most deservedly, that this second edition would
achieve far greater success than the first edition.
Dr. P.C.S. Krishnayya

Consultant, Chennai
and
Formerly Associate Professor,
IIT, Kharagpur
C HAPTER 29
Foreword
I am happy to introduce the revised edition of the book on

Transformers covering various aspects of design, manufacture, and
operation.
BHEL started manufacturing transformers in 1961 and has since
made significant strides. In this edition, efforts have been made to
reflect and put forward the expertise gained by BHEL in the field of
transformers over these forty years by practising engineers having
rich and vast experience. Certain new chapters have been added and
existing chapters have been updated as per the latest practices. For
the benefit of the readers, certain solved examples have also been
included at the end. I find the work informative and exhaustive as it
follows a practical approach, lacking in most of the other books on
the subject.
The transformer industry is now more than a century old. Signifi-
cant advances have been made and growth in technology continues
at a rapid pace. I sincerely hope that this book will serve the need of
a useful reference book not only among practising engineers but also
among the academicians.
K.G. Ramachandran
CMD–BHEL
New Delhi
C HAPTER 29
Preface to the Second Edition
Transformers was first published in the year 1987. With the publica-
tion of this second edition, the book celebrates its 15th anniversary.
The success of the book is evident from the increase in its demand
among practising engineers and academicians. Starting from the
fundamentals of design, manufacturing technology, materials used,
etc. the book caters to the latest developments, too. The chapters
provide an in depth knowledge and understanding of the intricacies
of transformers, including the various aspects of latest technological
upgradation. Six new chapters have been introduced in this edition
of the book:
§ Designing and manufacturing—A short-circuit proof trans-
former
§ High voltage condenser bushings
§ Computerization—A tool to enhance engineering prod-
uctivity
§ Condition monitoring, residual life assessment and
refurbishment of transformer
§ Controlled shunt reactor
§ An overview of transformer
Solved examples have also been included in the book to illustrate
better the design aspects of a power transformer.
With these enhancements, the book will serve as a useful
reference book for practising engineers engaged in design, manu-
facturing, planning, testing, erection, operation and maintenance of
transformers as well as for students and researchers.
This book is a collective effort of experts in the field of
transformers at Bharat Heavy Electricals Limited (BHEL), a leading
transformer manufacturer in India. My thanks are due to all the
authors who contributed the various chapters to this book. The
xiv Preface to the Second Edition
editorial committee headed by Shri T. K. Ganguli, an eminent

transformer designer, went over the complete material to integrate
it into a homogeneous book and ensure that all relevant aspects as
per the latest practices have been included in the revised edition. To
integrate the entire material and present it in a standard format was
an uphill task. Our grateful thanks are due to the editorial
committee members who put in exemplary efforts to lend this book
its present shape.
S. N. Roy
Executive Director
BHEL, Bhopal
C HAPTER 29
Preface to the First Edition
The book covers all the aspects of power transformers—basic funda-

mentals of the theory, design principles, manufacturing technology,
materials specifications and standards. It also covers special trans-
formers including reactors and inverter transformers for HVDC sys-
tems. It will serve as a reference book for practising engineers en-
gaged in the design, manufacture, planning, testing, erection, opera-
tion and maintenance of transformers. It will also be useful to stu-
dents and researchers.
This book is the collective effort of nineteen experts from Bharat
Heavy Electricals Ltd. (BHEL), a leading transformer manufacturer
in India. An editorial committee went over the complete material to
integrate it into a homogenous book and to ensure complete continu-
ity between the chapters.
Bharat Heavy Electricals Ltd. (BHEL), one of the largest engi-
neering organizations in the country, is engaged in the design and
supply of heavy equipment and total systems for generation, trans-
portation and utilization of energy.
R.K. Saraf
Executive Director
BHEL, Bhopal
C HAPTER 29
Acknowledgements
The editorial committee was given the task of bringing in the revised
edition of the book on Transformers. It is indeed commendable that
the committee could accomplish the task in a very short time. We
acknowledge with thanks the permission given by the management
of Bharat Heavy Electricals Ltd. for publishing this new edition.
Thanks are also due to all the authors who contributed chapters
to this book.
Integrating the entire material and presenting it as a homogenous
whole and at the same time, selecting new chapters and updating the
existing chapters to cover many other aspects in detail was an uphill
task. Our grateful thanks are also due to the editorial committee
members who put in exemplary efforts without which this book
could not have taken this shape.
P.T. Deo
Executive Director
BHEL, Delhi
Editorial Committee
Shri T.K. Ganguli
Shri M.V. Prabhakar
Shri S.K. Gupta
Shri R.K. Singh
Shri K. Gopal Krishnan
C HAPTER 29
Contents
Message I vii
Message II ix
Foreword xi
Preface to the Second Edition xiii
Preface to the First Edition xv
Acknowledgements xvii
Editorial Committee xviii
1. Introduction 1
• T.K. Ganguli
2. Principles of Transformers 6
• R.K. Singh
2.1 Working Principle of a Transformer 6
2.2 Elementary Theory of an Ideal Transformer 7
2.3 EMF Equation of a Transformer 8
2.4 Voltage Transformation Ratio (K) 9
2.5 Ideal Transformer 10
2.6 Transformer Having Winding Resistance
But No Magnetic Leakage 13
2.7 Magnetic Leakage 14
2.8 Transformer with Resistance and
Leakage Reactance 15
2.9 Equivalent Circuit 16
2.10 Regulation 21
2.11 Losses in Transformers 21
2.12 Efficiency 22
2.13 The Auto-Transformer 23
2.14 Three-winding Transformer 25
2.15 Parallel Operations of Transformers 27
Appendix 32
References 33
xx Contents
3. Materials Used in Transformers 34

• M.P. Singh, T.K. Ganguli
3.1 Insulating Oil 34
3.2 Insulating Paper 47
3.3 Pressboard 51
3.4 Wood 58
3.5 Insulated Copper Conductor for Windings 59
3.6 Crepe Paper Covered Flexible Copper Cable 66
3.7 Sealing Materials 66
3.8 Cold Rolled Grain-oriented Electrical Steel Sheet
(CRGO) 68
3.9 Structural Steel 81
3.10 Future Trends 84
References 85
4. Magnetic Circuit 86
• K.N. Labh, R.C. Agarwal
4.1 Material 86
4.2 Design of Magnetic Circuit 87
4.3 Optimum Design of Core 91
4.4 Manufacturing 99
References 107
5. Windings and Insulation 108
• M.V. Prabhakar, S.K. Gupta
5.1 Types of Windings 108
5.2 Surge Voltage Behaviour of Windings 116
5.3 Internal Heat Transfer in Windings 125
5.4 Insulation Design 126
5.5 Electric Field Plotting 130
5.6 Finite Difference Method (FDM) 131
References 136
6. Voltage Regulation and Tapchanger 138
• B.L. Rawat, A.K. Ekka
6.1 Off-circuit Tapchanger 138
6.2 On-load Tapchanger (OLTC) 141
6.3 Constructional and Operational
Features of OLTC 148
6.4 Manual and Electrical Operation of Tapchanger 150
6.5 Automatic Control of Tapchanger 151
6.6 Tapchanger Selection 151
6.7 Latest Trends in Tapchanger Design 152
Reference 153
Contents xxi
7. Electromagnetic Forces in Power Transformers 154

• M.V. Prabhakar, T.K. Ganguli
7.1 Leakage Flux in a Typical Two Winding
Transformer 154
7.2 Nature of Forces 154
7.3 Basic Formula and Methods for Force Evaluation 158
7.4 Radial Forces 159
7.5 Axial Forces 160
7.6 Roth’s Method of Force Calculation 163
7.7 Modes of Failure of Windings and Design
of Windings 166
7.8 Strengthening of Coils to Withstand Short-circuit
Forces 168
7.9 Design of Clamping Structures 169
7.10 List of Symbols 174
References 176
8. Cooling Arrangements 177
• C.M. Sharma
8.1 Various Types of Cooling 177
8.2 Cooling Arrangements 179
8.3 Propeller Type Fans 185
8.4 Transformer Oil Pump 188
8.5 Flow Indicators 190
8.6 Stress and Hydraulic Analysis of Pipework 191
9. Design Procedure 198
• R.C. Agarwal
9.1 Specifications of a Transformer 198
9.2 Selection of Core Diameter 198
9.3 Selection of Flux Density 199
9.4 Selection of Type of Core 200
9.5 Selection of Leg Length 200
9.6 Selection of Type of Windings 201
9.7 Selection of Tapchanger 202
9.8 Calculation of Number of Turns 203
9.9 Selection of Conductor and Current Density 203
9.11 Calculation of Lateral and Axial Dimensions
of Coils 204
9.12 Ampere-turn Balancing 205
9.13 Reactance Calculation 205
9.14 Iron Weight and Losses 210
9.15 Copper Weight and Losses 210
xxii Contents
9.16 Stray Losses in Transformer 211

9.17 Stray Loss Control 214
9.18 Impulse Calculation 218
9.19 Mechanical Forces in Windings 218
9.20 Temperature Gradient and Cooling Calculation 219
9.21 Typical Design Calculations for Two Winding
and Auto-transformers 219
References 226
10. Structural Design of Transformer Tank 227
• M.K. Shakya, S.G. Bokade
10.1 Types of Tank Constructions 227
10.2 Structural Design of Transformer Tanks 230
10.3 Testing of Tanks 244
References 250
11. Transformer Auxiliaries and
Oil Preservation Systems 251
• S.C. Verma, J.S. Kuntia
Transformer Auxiliaries 251
11.1 Gas Operated (Buchholz) Relay 251
11.2 Temperature Indicators 253
11.3 Pressure Relief Valve 255
11.4 Oil Level Indicator 256
11.5 Bushing and Cable Sealing Box 257
Transformer Oil Preservation Systems 258
11.6 Silicagel Breather 260
11.7 Gas Sealed Conservators 260
11.8 Thermosyphon Filters 263
11.9 Bellows and Diaphragm Sealed Conservators 264
11.10 Refrigeration Breathers 265
12. Manufacturing and Assembly 267
• T.K. Ganguli, M.V. Prabhakar
12.1 Core Building 267
12.2 Preparation of Windings 268
12.3 Winding Assembly 269
12.4 Core and Winding Assembly 271
12.5 Terminal Gear Assembly 271
12.6 Placement of Core and Winding
Assembly in Tank 272
12.7 Processing 273
12.8 Servicing of Transformer 273
12.9 Tanking 274
Contents xxiii
13. Drying and Impregnation 275

• M.P. Singh, M.V. Prabhakar
13.1 Basic Principles of Drying 276
Principles of Drying 277
13.2 Conventional Vacuum Drying 278
13.3 Vapour Phase Drying 281
References 290
14. Testing of Transformers and Reactors 291
• P.C. Mahajan, M.L. Jain, R.K. Tiwari
Section I 292
14.1 Testing of Power Transformers 292
Section II 305
14.2 Impulse Testing 305
References 320
Section III 321
14.3 Partial Discharge Testing 321
References 334
Section IV 335
14.4 Testing of Reactors 335
References 349
Section V 350
14.5 Short Circuit Testing of Power Transformers 350
15. Standards on Power Transformers 358
• V.K. Lakhiani, S.K. Mahajan
15.1 First Revision of IS: 2026 359
15.2 Other Related Standards 364
15.3 New Standards 367
15.4 Standard Specification of a Power Transformer 368
Appendices 372
16. Loading and Life of Transformers 377
• D.P. Gupta
16.1 Life of a Transformer 378
16.2 Ageing of Insulation 379
16.3 Law of Insulation Ageing 381
16.4 Significance of Weighted Value of
Ambient Temperature 384
16.5 Relationship between Weighted Ambient,
Winding Rise and Hot-spot Temperatures 386
16.6 Determination of Weighted
Ambient Temperature 394
16.7 Relationship between Weighted Ambient
and Load 399
xxiv Contents
16.8 Alternative Approach for the Calculation

of Weighted Ambient 400
16.9 Transformer Loading Guides 407
16.10 Loading by Hot-Spot Temperature Measured
by WTI 409
16.11 Selection and Use of a Transformer 411
References 413
Annexures 414
17. Erection and Commissioning 439
• C.M. Shrivastava
17.1 Despatch 439
17.2 Inspection upon Arrival at Site 440
17.3 Handling 440
17.4 Installation 441
17.5 Commissioning 447
17.6 Maintenance 449
17.7 Dos for Power Transformer 451
17.8 Don’ts for Power Transformers 453
17.9 Dos and Don’ts for HV Condenser Bushings 454
18. Transformer Protection 456
• B.L. Rawat
18.1 Protection against External Faults 456
18.2 Protection against Internal Faults 459
References 465
19. Reactors 466
• C.M. Shrivastava, S.K. Mahajan
19.1 Series Reactors 466
19.2 Shunt Reactor 473
19.3 Neutral Earthing Reactor 482
19.4 Tuning for Filter Reactors 482
19.5 Arc Suppression Reactors 483
19.6 Earthing Transformers (Neutral Couplers) 483
19.7 Standards on Reactors 484
20. Traction Transformers 486
• J.M. Malik
20.1 Types of Traction Transformers 486
20.2 Special Considerations 487
20.3 Design and Constructional Features 489
20.4 Traction Transformer for Thyristor
Controlled Locomotives 493
References 495
Contents xxv
21. Rectifier Transformers 496

• J.S. Sastry
21.1 Comparison between Rectifier Transformer
and Power Transformer 497
21.2 Rectifier Circuits 499
21.3 Design Features of Rectifier Transformers 503
21.4 Transductors 517
21.5 Constructional Features of
Rectifier Transformers 518
21.6 Tank Design 521
21.7 Testing 523
References 525
22. Convertor Transformers for HVDC Systems 526
• I.C. Tayal, C.M. Sharma, S.C. Bhageria
22.2 Higher Harmonic Currents 529
22.3 DC Magnetization 529
22.4 DC Bushings 530
22.5 On Load Tapchanger 530
22.6 Influence of Impedance Variation 530
22.7 Connections 531
22.8 Specifications 533
22.9 Manufacturing Features 533
22.10 Tests 536
References 538
23. Controlled Shunt Reactor 539
• S.C. Bhageria, J.S. Kuntia
23.1 Controlled Shunt Reactor (CSR) Principle 539
23.2 Controlled Shunt Reactor Transformer (CSRT) 541
23.3 Special Features of the Controlled
Shunt Reactor Transformer 545
23.4 Controlled Shunt Reactor Much More
Than a Shunt Reactor 549
23.5 Conclusion 550
References 550
24. Designing and Manufacturing —A Short
Circuit Proof Transformer 551
• T.K. Ganguli, S.K. Gupta
24.1 Forces During Short-Circuit 552
24.2 Various Considerations to Design a Transformer
Suitable for Short-Circuit Duty 556
24.3 Manufacturing Aspects 559
xxvi Contents
24.4 Quality Aspects 560

24.5 Conclusion 560
25. High Voltage Condenser Bushings 562
• R.K. Agarwal, Assem Dhamija
25.1 Introduction 562
25.2 Classification of Bushings 562
25.3 Design of Bushing 565
25.4 Constructional Details and Main Parts
of Bushing 567
25.5 Testing of Bushing 571
25.6 Factors Affecting the Performance
of Bushing 573
25.7 Condition Monitoring at Site 574
25.8 Dos and Don’ts for HV Condenser Bushings 575
26. Computerisation—A Tool to Enhance Engineering
Productivity 577
• R. Mitra, DGM/TRE
27. Condition Monitoring, Residual Life
Assessment and Refurbishment of
Transformers 581
• C.M. Shrivastava, T.S.R. Murthy
Section A 581
27.1 Analysis Method 582
27.2 Physical Inspection 584
27.3 Residual Life Assessment (RLA) 588
27.4 Different Techniques for Life Estimation 589
27.5 Methodology Adopted 591
27.6 Conventional and Special Tests on Oil Samples 592
27.7 Refurbishment 593
27.8 Conclusion 596
Annexure I 597
28. Transformers: An Overview 598
• S.N. Roy, P.T. Deo
Solved Examples 603
Index 609
C HAPTER 1
Introduction
T.K. Ganguli
The history of transformer goes back to the early 1880s. With the
sharp increase in demand for electric power, power transformers in
400 kV ratings were produced as early as 1950. In the early 1970s
unit ratings as large as 1100 MVA were produced and 800 kV and
even higher kV class transformers were manufactured in the early
1980s.
A transformer is a static piece of equipment with a complicated
electromagnetic circuit inside. The energy is transferred from one
electrical circuit to another through the magnetic field. In its sim-
plest form, a transformer consists of two conducting coils having a
mutual inductance. In an ideal case it is assumed that all the flux
linked with the primary winding also links the secondary winding.
But, in practice it is impossible to realize this condition as magnetic
flux cannot be confined. The greater portion of the flux flows in the
core while a small portion called the leakage flux links one or the
other winding. Depending upon the particular application and type
of connection, a transformer may have additional windings apart
from the two conventional windings. Chapter 2 deals with the prin-
ciples of transformers mainly covering the basic electromagnetic
force (emf) equation, ideal transformer, equivalent circuit, calcula-
tion of regulation, no load and load losses, efficiency and parallel
operation of transformers.
Major materials like copper, cold-rolled grain oriented silicon
steel, insulating oil, pressboard and paper insulation and certain
ferrous and nonferrous items are essential to build a compact and
trouble-free transformer. Chapter 3 has been primarily devoted to
explain the characteristics of these materials which shall be helpful
in selecting the correct material for the equipment.
2 Transformers, 2/e
Designing an insulation system for application in higher voltage

class transformer is an art and with the use of the best materials
available today, it is possible to economies on size as well as produce
a reliable piece of equipment. Keeping in view the transportability,
operational limitations and guaranteed technical performance of
the transformer, particular type of core construction is adopted.
Chapter 4 is devoted to design of magnetic circuit, constructional
features, manufacturing, assembly and finally fitting core in the
tank.
Windings and insulation arrangement has been covered in Chap-
ter 5. Starting from spiral and layer type of windings, more empha-
sis is given to helical and disc types which are more commonly used
in the latest design practices. The impulse voltage withstand behav-
ior of an ordinary disc winding can be enhanced by interleaving the
disc winding. The initial voltage distribution at the line end should
be taken care of for high voltage windings. The transient voltage
distribution and internal heat transfer in the windings have been
described in detail. The insulation design becomes more complex as
we move towards higher and higher voltage class of transformer
windings. Proper sizing and routing can be further examined by de-
tailed electrostatic field plots. To vary the voltage in a transformer,
tap changers are used which have a different type of regulating
winding connection, viz. linear, reversing, coarse fine, etc. Chapter
6 describes the tap changer types, constructional features and its
control.
Designing a transformer to cater to electromagnetic forces has
been covered in Chapter 7. Radial and axial forces occurring during
short circuit or line faults can be calculated and winding design can
be finalized to withstand these forces. The clamping structures put
the coils under a pressure, higher than that produced by short cir-
cuit forces. An effort has been made to describe the method for cal-
culating stresses and then dimensioning the clamping structures
adequately. The cooling of a transformer becomes more relevant
from the point of view of ageing of insulation system and ensuring
longer life due to less thermal degradation. Chapter 8 covers vari-
ous types of cooling, its arrangement and design calculation.
The design procedure described in Chapter 9 gives every detail of
a transformer starting from the selection of core-size, winding con-
ductor, reactance calculation and then realization of main guaran-
Introduction 3
teed parameters like percentage impedance, no load loss, load loss,

etc. Estimation and control of stray losses, winding gradient has
also been described. Typical design calculation for two winding and
auto transformers has been covered in brief. Structural design is
described in Chapter 10 which gives an idea of stresses that are
developed in main tank and other supporting structures.
Proper selection of transformer auxiliaries is essential for ensur-
ing safe operation of the main equipment and provides protection
under fault conditions. Chapter 11 covers the major auxiliaries like
gas operated buchholz relay, temperature indicators, pressure relief
valve, bushings, cable box, oil preservation system, etc. Assembly
and manufacturing aspects have been covered in Chapter 12. After
the winding and core assembly is completed, drying and impregna-
tion is a vital process that is described in Chapter 13. Cellulose in-
sulation used in power transformers and reactors has approxi-
mately 6 to 10% of moisture by weight at ambient temperature, be-
ing a hygroscopic material. Vacuum drying and vapour phase dry-
ing is an important tool to extract moisture from the insulation
items. Uniform heating of the entire active part (core and winding)
mass must be ensured so that shrinkage and moisture extraction is
optimum.
To ensure quality and conformation to design calculations, test-
ing is an important activity in the transformer. The basic testing
requirement and testing codes are set out in national and interna-
tional standards. Chapter 14 is intended to cover the purpose and
methodology of performing the tests. Chapter 15 covers the various
standards generally used for transformers. Depending upon the
consumer requirement, sometimes the transformers are loaded be-
yond its nameplate rating for a brief period. Chapter 16 covers the
loading and life of a transformer.
Chapter 17 describes some of the main precautions which must
be taken during erection and commissioning of a transformer.
Maintenance schedule has also been discussed so that continuous
trouble free service could be ensured. Inspection upon arrival at
site, installation, oil filling, drying of transformer and analysis of
gases of power transformers, etc. have been explained for the ben-
efit of the users. The transformer being a vital equipment, its pro-
tection is equally important. Some of the basic protection schemes
have been dealt with in Chapter 18. Reactors are usually classified
4 Transformers, 2/e
according to duty application, viz. current limiting, neutral

earthing, shunt, smoothing, etc. These reactors have typical charac-
teristic requirements and call for different constructions, viz. with
air core or with gapped iron core for fixed and variable reactance.
Chapter 19 deals in detail with the design, and construction etc. of
different types of reactors. Chapters 20 and 21 mainly deal with
special types of transformers. For chemical plants, aluminium
plants, etc. where electrolytic processes are adopted and the direct
current requirement is quite large, rectifier transformers are used.
Traction transformers require special considerations for their de-
sign due to a limitation of space availability and problems due to
vibration. They have to be designed and manufactured to stringent
specifications, so as to withstand heavy stresses in this type of ap-
plication.
The growth of HVDC system for transporting power to long dis-
tances calls for discussing transformers for such application, too.
Chapter 22 covers the basic aspects of a converter transformer,
which is one of the most expensive equipment in HVDC system.
Controlled shunt reactors can be employed in place of fixed shunt
reactors and Chapter 23 is fully devoted to this new area. The ad-
vantages of controlled shunt reactors are mainly fully controllable
reactive power, reduced dynamic over voltages, increased power
carrying capacity of lines, faster responses etc. Chapter 24 describes
the criteria for designing and manufacturing a short circuit proof
transformer. Steady increase in unit ratings of transformers and
simultaneous growth of short circuit levels of network have made
the short circuit withstand capability of the transformer one of the
key aspects of its design. Additionally, aspects like material selec-
tion, selection of fittings and manufacturing processes, etc. are to be
carefully examined in detail. Chapter 25 covering high voltage con-
denser bushing has been specially designed to cover application
guidelines, use of insulating material, design, construction and test-
ing, etc. The factors affecting the performance of bushing and pre-
ventive measures like condition monitoring, etc. have been dealt
with in detail. In the present day scenario, it is not possible to func-
tion without the application of computer aided design. Accordingly
Chapter 26 gives an idea for catering to the high demand for tailor-
made transformer products and the ways to overcome this problem.
Introduction 5
The need for a reliable and stable system is being increasingly

felt. Thus Chapter 27 is devoted to condition monitoring, residual
life assessment and refurbishment of the transformer. The residual
life assessment study is carried out to predict the health of the
transformer insulation and remaining life of the transformer. Cellu-
lose insulation degrades due to heating or electrical breakdown, re-
sulting in the production of furfural derivatives which dissolve in
oil. Hence, chemical analysis of the oil gives evidence of changes
that are taking place in the winding insulation during operation.
Due to paucity of funds, it is not practical and economical to replace
old units with new ones. In such cases, refurbishing/retrofitting is
an economical and viable alternative. Finally, Chapter 28 provides
an overview of the transformer and describes in general the salient
points which contribute an essential part in producing the equip-
ment. Although the transformer is a complex piece of equipment
mainly produced manually, efforts on various fronts, viz. design,
manufacture, vendor development, process mapping and maintain-
ing history card, etc. contribute as important tools for producing a
reliable piece of transformer.
This book is an attempt to include in one volume the various ma-
jor aspects related to a power transformer. An effort has been made
to give a comprehensive and up-to-date coverage of the latest tech-
nologies and developments in the field of power transformers. Fi-
nally, some solved examples have also been included at the end of
the book to aid better understanding especially for students and
practising engineers. The topics covered in the book shall be helpful
for design, manufacture, testing, erection, commissioning and
maintenance of power transformers in every corner of the world.
C HAPTER 2
Principles
of Transformers
R.K. Singh
2.1 Working Principle of a Transformer
A transformer is a static piece of apparatus used for transferring

power from one circuit to another without change in frequency. It
can raise or lower the voltage with a corresponding decrease or
increase in current. In its simplest form, a transformer consists of
two conducting coils having a mutual inductance. The primary is
the winding which receives electric power, and the secondary is the
one which may deliver it. The coils are wound on a laminated core of
magnetic material.
The physical basis of a transformer is mutual inductance between
two circuits linked by a common magnetic flux through a path of low
reluctance as shown in Fig. 2.1.
The two coils possess high mutual inductance. If one coil is
connected to a source of alternating voltage, an alternating flux is
set up in the laminated core, most of which is linked up with the
other coil in which it produces mutually induced emf (electromotive
force) according to Faraday’s laws electromagnetic induction, i.e.
di
e =M
dt
where, e = induced emf
M = mutual inductance
If the second circuit is closed, a current flows in it and so electric
energy is transferred (entirely magnetically) from the first coil
(primary winding) to the second coil (secondary winding).
Principles of Transformers 7
Laminated core
Secondary
Primary
Figure 2.1 Diagram showing magnetic circuit and windings of

a transformer.
2.2 Elementary Theory of an Ideal Transformer
An ideal transformer is one which has no losses, i.e. its windings

have no ohmic resistance and there is no magnetic leakage. In other
words, an ideal transformer consists of two coils which are purely
inductive and wound on a loss-free core.
It may, however, be noted that it is impossible to realize such a
transformer in practice, yet for convenience, we will first analyse
such a transformer and then an actual transformer.
Consider an ideal transformer [Fig. 2.2(a)] whose secondary is
open and whose primary is connected to a sinusoidal alternating
voltage v1. Under this condition, the primary draws current from
the source to build up a counter electromotive force equal and oppo-
site to the applied voltage.
Since the primary coil is purely inductive and there is no output,
the primary draws the magnetising current Im only. The function of
this current is merely to magnetise the core, it is small in magnitude
and lags v1 by 90°. This alternating current Im produces an alternat-
ing flux f which is proportional to the current and hence is in phase
with it. This changing flux is linked with both the windings. There-
8 Transformers, 2/e
f V1
90° Im
V1 E1 E2 O
V2 90° f
E1
V2 = E2
(a) (b)
Figure 2.2 Vectorial representation of induced emf of an ideal

transformer under no-load.
fore, it produces self-induced emf in the primary. This self-induced

emf e1 is, at any instant, equal to and in opposition to v1. It is also
known as counter emf of the primary.
Similarly in the secondary winding, an induced emf e2 is pro-
duced which is known as mutually induced emf. The emf is in phase
opposition with v1 and its magnitude is proportional to the rate of
change of flux and the number of secondary turns. Figure 2.2(b)
shows the vectorial representations of the above quantities.
2.3 EMF Equation of a Transformer
Let N1 = Number of turns in primary

N2 = Number of turns in secondary
fm = Maximum flux in the core in webers
Bm = Flux density in weber/sq m (Tesla)
A= Net cross-sectional area of core in sq m
f= Frequency of ac input in Hz
v1 = Instantaneous value of applied voltage in primary
winding in volts
v1m = Maximum value of applied voltage in volts.
The instantaneous value of counter electromotive force e1 is

df
e 1 = – N1 volt (2.1)
dt
As discussed above in Sec. (2.2) the counter emf e1 is equal and
opposite to applied voltage v1, i.e.
df
v 1 = N1
dt
If the applied voltage is sinusoidal, that is
v1 = v1m sin 2p ft
Then f = fm sin 2p ft
Hence e1 = – N1 fm cos 2p ft ¥ 2p f
These equations are expressed as vectors as shown in Fig. 2.2(b),
where V1 and E1 are the rms values of v1 and e1. To obtain the rms
value of counter emf e1, divide its maximum value given above
by 2 .
2p
Then E1 = f N1 f m (2.1.1)
2
The cosine term has no significance except to derive the instanta-
neous values.
i.e. E1 = 4.44 f N1 fm
or E1 = 4.44 f N1 Bm A (2.2)
Similarly rms value of emf induced in secondary is,
E2 = 4.44 f N2 Bm A (2.3)
In an ideal transformer
V1 = E1
and V2 = E2
where V2 is the secondary terminal voltage [Fig. 2.2(b)].
2.4 Voltage Transformation Ratio (K)
From Eqs. (2.2) and (2.3), we get

E2 N
= 2 =K (2.4)
E1 N1
This constant is known as voltage transformation ratio.
(a) If N2 > N1, i.e., K > 1, then the transformation is called as
step-up transformer.
10 Transformers, 2/e
(b) If N2 < N1, i.e., K < 1, then the transformer is called as step-
down transformer.
Again for an ideal transformer
Input = Output
V1I1 = V2I2 (neglecting Im)
I2 V 1
or = 1 = (2.5)
I1 V2 K
where I1 and I2 are primary and secondary currents.
Hence the currents are in the inverse ratio of the transformation
ratio.
2.5 Ideal Transformer
We will consider two cases,

(a) when such a transformer is on no-load and
(b) when it is loaded.
(a) Transformer on No-load

The primary input current under no-load condition has to supply
(i) iron-loss in the core i.e., hysteresis loss and eddy current loss and
(ii) a very small amount of copper-loss in primary. Hence the no-load
primary input current I0 is not at 90° behind v1 but lags it by an
angle q0 which is less than 90°. No-load primary input power W0 =
V1 I0 cos q0. No-load condition of an actual transformer is shown
vectorially in Fig. 2.3.
As seen from Fig. 2.3, primary current I0 has two components.
(i) One in phase with V1. This is known as active or working or
iron-loss component Iw, because it supplies the iron-loss plus
a small quantity of primary Cu-loss.
Iw = I0 cos q0 (2.6)
(ii) The other component is in quadrature with V1 and is known
as magnetizing component because its function is to sustain
the alternating flux in the core. It is wattless.
Im = I0 sin q0 (2.7)
Obviously I0 is the vector sum of Iw and Im, hence
I0 = (Im2 + Iw2 (2.8)

The no-load primary current I0 is very small as compared to full-
load primary current. As I0 is very small, hence no-load primary
V1
Iw I0
f
Im
E1
E2
Figure 2.3 No-load current of an ideal transformer (having core

losses only).
copper-loss is negligibly small which means that no-load primary

input is practically equal to the iron-loss in a transformer.
(b) Transformer on-load

When the secondary is loaded, secondary current I2 is set up. The
magnitude of I2 is determined by the characteristic of the load. The
secondary current sets up its own mmf (= N2 I2) and hence its own
flux f2 which is in opposition to the main primary f, which is due to
I0. The opposing secondary flux f2 weakens the primary flux mo-
mentarily and primary back emf E1 tends to reduce. For a moment
V1 gains the upper hand over E1 and hence causes more current (I¢2)
to flow in primary.
The current I¢2 is known as load component of primary current.
This current is in phase opposition to current I2. The additional
primary mmf N2 I¢2 sets up a flux f¢2 which opposes f2 (but is in the
same direction as f) and is equal to it in magnitude. Thus, the mag-
netic effects of secondary current I2 get neutralized immediately by
additional primary current I¢2. The whole process is illustrated in
Fig. 2.4. Hence, whatever may be the load conditions, the net flux
f
I0
V1 V2
f
I0 Load
V1 I2
f2
f
I 2¢ f2
Load
I0
V1 I2
f2
f
I1 Load
V1 I2
Figure 2.4 Transformer on-load.
passing through the core is approximately the same as at no-load.

Due to this reason the core-loss is also practically the same under
all load conditions.
As f2 = f¢2
N2I 2 = N1I¢2
N2
I¢2 = ¥ I2 = KI2
N1
Hence, when transformer is on load, the primary winding has two
currents I0 and I¢2 (which is antiphase with I2 and K times its magni-
tude). The total primary current is the vector sum of I0 and I¢2. In
Fig. 2.5 are shown the vector diagrams for a loaded transformer. In
Fig. 2.5(a), current I¢2 is in phase with E2 (for non-inductive loads).
In Fig. 2.5(b), it is lagging behind E2 (for inductive loads).
V1 V1
V1
I1
I1 I2¢ I2¢ = I1
q1
I2¢
q1
q1 I0
q0 I0
q q q
I2
q2 q2
I2
I2
E2 E2 E2
K=1 K=1 K=1
(I0 Neglected)
(a) (b) (c)
Figure 2.5 Vector diagram of an ideal transformer.
If we neglect I0 as compared to I¢2 as shown in Fig. 2.5(c), then f1 = f2

and thus
N1I2¢ = N1I1 = N2 I2
I1 N
= 2 =K
I2 N1
It shows that under load conditions, the ratio of primary and
secondary currents is constant.
2.6 Transformer Having Winding Resistance

But No Magnetic Leakage
An ideal transformer was supposed to possess no resistance but an

actual transformer has primary and secondary windings with some
resistances. Due to these resistances, there is some voltage drop in
the two windings. The result is that:

(a) The secondary terminal voltage V2 is equal to the vector
difference of the secondary induced emf E2 and I2 R2 where
R2 is the resistance of the secondary winding.
V2 = E2 – I2 R2 (2.9)
(b) Similarly primary induced emf E1 is equal to the vector
difference of V1 and I1 R1 where R1 is the resistance of the
primary winding.
E 1 = V1 – I 1 R 1 (2.10)
The vector diagrams for non-inductive, inductive and capacitive
loads are shown in Fig. 2.6(a), (b) and (c) respectively.
I1R1 V1 I1R1
V1
I1R1 V1
E1 E1 I ¢2 = KI2 E1
I1 I1
I1
I ¢2
I ¢2 = KI2
q1
I0 I0 I0
f
q f
q2 f2
I2 I2
V2 I2
V2 V2
I2R2 I2R2 E2
E2 E2 I2R2
K=1
K=1 K=1 I0 neglected
(a) (b) (c)
Figure 2.6 Vector diagram of an ideal transformer (having core

losses and winding resistance).
2.7 Magnetic Leakage
In an ideal case it is assumed that all the flux linked with the
primary winding also links the secondary winding. But, in practice
it is impossible to realize this condition as magnetic flux cannot be
confined. The greater portion of the flux (i.e., the mutual flux) flows
in the core while a small proportion (Fig. 2.7) called the leakage flux
links one or the other winding, but not both. On account of the
leakage flux, both the primary and secondary windings have leakage
reactance, that is, each will become the seat of an emf of self-induc-
tion, of a magnitude equal to a small fraction of the emf due to main
flux. The terminal voltage V1 applied to the primary must, therefore,
have a component I1 X1 (where X1 is leakage reactance of primary) to
balance the primary leakage emf. In the secondary, similarly, an emf
of self-induction I2 X2 (where X2 is leakage reactance of secondary) is
developed. The primary and secondary coils in Fig. 2.7 are shown on
Leakage flux Main flux
Figure 2.7 Leakage flux.
separate limbs, an arrangement that would result in an exception-

ally large leakage. Leakage between primary and secondary could
be eliminated if the windings could be made to occupy the same
space. This, of course, is physically impossible, but an approximation
to it is achieved if the coils of primary and secondary are placed
concentrically. Such an arrangement leads to a marked reduction of
the leakage reactance. If on the other hand, the primary secondary
are kept separate and widely spaced, there will be much room for
leakage flux and the leakage reactance will be greater.
2.8 Transformer with Resistance and Leakage Reactance
Figure 2.8 shows the primary and secondary windings of a

transformer with resistance and leakage reactances taken out of the
R1 X1 R2 X2
Z1 Z2
Figure 2.8 Resistances and leakage reactances.
windings. The primary impedance is given by
Z1 = ( R12 + X 12 )
And the secondary impedance is given by
Z2 = ( R22 + X 22 )
V1 = E1 + I1 (R1 + jX1) = E1 + I1 Z1 (2.11)
and E2 = V2 + I2 (R2 + jX2) = V2 + I2Z2 (2.12)
The vector diagram of such a transformer for different kinds of
loads is shown in Fig. 2.9. In these diagrams vectors for resistive
drops are drawn parallel to current vectors, whereas reactive drops
are perpendicular to the current vectors. The angle q1 between V1
and I1 gives the power-factor angle of the transformer.
2.9 Equivalent Circuit
The transformer shown in Fig. 2.10(a) can be represented by an

equivalent circuit in which the resistance and leakage reactance of
the transformer are imagined to be external to the winding. The no-
load current I0 is simulated by a pure inductance X0 taking the mag-
netizing component Im and a noninductive resistance R0 taking the
active component Iw connected in parallel across the primary circuit
as shown in Fig. 2.10(b).
To make the transformer calculations simpler, it is preferable to
transfer voltage, current and impedance to the primary side.
The primary equivalent of the secondary induced voltage is
E2¢ = E2/K = E1
(where K is the transformation ratio)
I1X1 I1X1 I1R1

V1 I1X1 V1
V1
I 1Z 1 I1R1 I 1Z 1 I1R1 I 1Z 1
E1 E1 E1
I1
I1
I1
I ¢2 I ¢2
q1 I ¢2 q
I0 I0 I0
f
f f
f2
f2 I2
I2
V2 V2
I 2Z 2 I 2Z 2
I2R2
I2R2 E2 E2 V2 I 2Z 2
I2X2 I2X2 E2
I2R2
I2
I2X2
Figure 2.9 Vector diagram of transformer.
Similarly, primary equivalent of secondary terminal or output

voltage is
V2¢ = V2/K
Primary equivalent of secondary current is
I2¢ = I2K
and R2¢ = R2/K2,
X2¢ = X2/K2,
Z2¢ = Z2/K2
The same relationship is used, for shifting an external load im-
pedance to primary. The secondary circuit is shown separately in
Fig. 2.11(a) and its equivalent primary values in Fig. 2.11(b).
The equivalent circuit with secondary parameters transferred on
primary side is given in Fig. 2.12. This is known as exact equivalent
I1 I2
E2
V1 E1 V2
(a)
Z1 Z2
X1 I0 R2 X2
R1
Im
Iw E1 E2 V1
V1
R0 ZL
X0
Ideal transformer
(b)
Figure 2.10 Parameters of transformer.
Z2
I2 R 2¢ = R2/K 2 X 2¢ = X2/K 2
R2 X2 I 2¢
2
ZL Z L¢ = ZL/K
E2 V2 Load E ¢2 = E1 V 2¢ = V2/K Load
(a) (b)
Figure 2.11 Secondary resistance and reactance referred to primary.
circuit. A simplification of the equivalent circuit can be further

made by transferring the exciting circuit across the input terminals
as shown in Fig. 2.13 or in Fig. 2.14.
It should be noted that in this case R0 = V1/Iw and X0 = V1/Im.
Further simplification may be achieved by omitting I0 altogether as
shown in Fig. 2.15(a).
Equivalent circuit for a three-winding transformer described in
cl. 2.14 is given in Fig. 2.15(b) as a star network.
Z1 Z 2¢
I1
R1 X1
I0 I 2¢ R 2¢ X 2¢
Iw Im
V1 E ¢2 = E1 V 2¢ Z L¢
R0 X0
I1 I0 I 2¢
Figure 2.12 Exact equivalent circuit.
I1 I 2¢ I 2¢
R1 X1 R 2¢ X 2¢
I0
Iw Im
V1 E ¢2 = E1 Z L¢
R0 X0
I1 I0 I 2¢
Figure 2.13 Simplified equivalent circuit.
I1 R1 R 2¢ X1 X 2¢
I 2¢
I0 R01 X01
Iw Im
Z L¢
V1 V 2¢
R0 X0
I0 I2
I1
Figure 2.14 Simplified equivalent circuit.

Z01
I1 = I ¢2
R01 X01
Z L¢
V1 V 2¢
(a)
2
R2 X2
R1 X1
3
R3 X3
(b)
Figure 2.15 (a) Simplified equivalent circuit.

(b) Equivalent circuit for three-winding transformer.
From Fig. 2.12 it is found that total impedance between the input
terminals is
LM OP
1
Z = Mz 1 + P (2.13)
MM 1 1 P
+
N z ¢ + z ¢ z PQ
2 1 m
where Z2¢ = R2¢ + jX2¢
Zm = impedance of the exciting circuit
= R0 + jX0
LM OP
1
V 1 = I1 MMZ 1 +
1 1 P
P (2.14)
+
MN Z ¢ + Z ¢ Z PQ
2 L m
2.10 Regulation
When the transformer is loaded, with a constant primary voltage,

then the secondary terminal voltage drops (assuming lagging
power-factor); it will increase if power factor is leading, because of
its internal resistance and leakage reactance.
Let V2 = secondary terminal voltage at no-load
V2¢ = secondary terminal voltage at load
% Regulation =
LM V - V ¢ OP ¥ 100
2 2
N V Q2
Then, % regulation of a loaded transformer at any power factor is
given as
( X cos q - R sin q ) 2
= (R cos q + X sin q) + (2.15)
200
where R = percentage resistive drop
X = percentage reactive drop
cos q = lagging power-factor
It is to be noted that in case of leading power-factor, q will change
to –q.
Typical calculations for computation of percentage regulation at
different load and power factor are given in Appendix 2.1.
2.11 Losses in Transformers
Losses in transformers are
(a) No-load Losses

It includes both hysteresis loss and eddy current loss. As the core
flux in a transformer remains practically constant at all loads, the
core-loss is also constant at all loads.
Hysteresis loss Wh = Khf Bm1.6 watts (2.16)
Eddy current loss We = Ke f 2 Kf 2 Bm2 watts (2.17)
where Kh = the hysteresis constant
Ke = the eddy current constant
Kf = the form factor
These losses are minimized by using steel of high silicon content
for the core and by using very thin laminations. The input power of
a transformer, when on no-load, measures the core-loss.
(b) Load Losses

This loss is mainly due to the ohmic resistance of the transformer
winding. Copper-loss also includes the stray loss occurring in the
mechanical structure and winding conductor due to the stray fluxes.
Copper-loss (I2R-loss plus stray loss) is measured by the short-cir-
cuit test.
2.12 Efficiency
Output
% Efficiency = ¥ 100
Input
(Input - Losses)
= ¥ 100
Input
= 1-
FG Losses
¥ 100
IJ
H Input K
2.12.1 Condition of Maximum Efficiency
Copper-Loss Wc = I12 R01 = I22 R02
where R01 and R02 are equivalent resistance referred to primary and
secondary sides respectively.
Iron-Loss Wi = Wh + We
Primary input = V1I1 cos q1
LM (I R - W ) OP
Efficiency h = 1 -
2
1 01 i
MN V I cos q PQ i 1 1
L I R - W OP
= M1 - 1 01 i
(2.18)
N V cos q V I cos q Q
1 1 1 1 1
Differentiating both sides with respect to I1, we get

dh
=0–
R01F +
Wi I F I
dI1
GH
V1 cos q 1 2
V1 I1 cos q 1
JK GH JK
for h to be maximum,
dh
=0
dI1
hence the above equation becomes

R01 Wi
= 2
V1 cos q 1 V1 I1 cos q 1
or Wi = I12 R01
or = I22 R02
or Iron-loss = Copper-loss
i.e., maximum efficiency occurs when iron-loss is equal to copper-
loss.
The load corresponding to maximum efficiency is given by
=
FG Iron-loss IJ ¥ Full-load
H Full-load copper-loss K
Typical calculations for computation of percentage efficiency at
different loads and power factors are given in Appendix 2.1.
2.13 The Auto-Transformer
The auto-transformer has a single continuous winding which is

used for the input and output voltages, as shown in the Fig. 2.16. A
portion BC of the primary winding AB is used as secondary winding.
It is used where transformation ratio differs a little from unity. Its
theory and operation principles are similar to that of a two-winding
transformer. If N1 is the primary winding turns and N2 (a portion of
N1) is secondary winding turns, the transformation ratio K can be
represented in the same way as a two-winding transformer
V2 N
= 2 =K
V1 N1
The current in the secondary winding (may be called as common
winding) is the vector difference of I2 and I1. But as the two currents
are practically in phase opposition, the common winding current
may be taken as arithmetical difference of I2 and I1 i.e. (I2 – I1)
where I2 is greater than I1.
2.13.1 Savings of Copper in Auto-transformers

With reference to Fig. 2.16.
Weight of copper in AC section (series winding) is proportional to
(N1 – N2)I1
I1
A
C I2
V1
I2 – I1
V2 Load
I1 B I2
Figure 2.16 Auto-transformer.
Weight of copper in BC section (common winding) is proportional to

N2(I2 – I1)
So, total weight of copper in auto-transformer is proportional to
(N1 – N2) I1 + N2(I2 – I1)
In a two-winding transformer,
total weight of copper is proportional to (N1I1 + N2I2)
Weight of copper in auto-transformer
=
Weight of copper in two-winding transformer
( N1 - N2 ) I1 + N 2 ( I2 - I1 )
=
N1 I1 + N2 I2
N2
2¥
N1
=1–
FG 1 + N ¥ I IJ 2 2
H N IK 1 1
=1–G
F 2 K IJ = (1 – K)
H2K
FG where N2
=
I1
= KJ
I
H N 1 I2 K
or weight of copper in auto-transformer (Wa )
= (1 – K) ¥ [weight of copper in ordinary two-

winding transformer (W0)]
Hence, saving = W0 – Wa = W0 – (1 – K)W0 = KW0
or Saving = Transformation ratio (K) ¥ weight of copper in
ordinary two-winding transformer
Thus, it can be visualized that saving will increase as K approaches
unity.
2.14 Three-winding Transformer
2.14.1 Purpose
A transformer may have additional windings apart from the two
conventional main windings depending upon the particular applica-
tion and type of connection (of the main windings). In three-winding
transformers, the third winding is normally called as tertiary wind-
ing and it is provided to meet one more of the following require-
ments:
(a) For an additional load which for some reason must be kept
isolated from that of secondary.
(b) To supply phase-compensating devices, such as condensers,
operated at some voltage not equal to primary or secondary
or with some different connection (e.g. mesh).
(c) In star/star-connected transformers, to allow sufficient earth
fault current (zero-sequence component current) to flow for
operation of protective gear, to suppress harmonic voltages
and to limit voltage unbalance when the main load is asym-
metrical, the tertiary winding is delta-connected.
(d) As a voltage coil in a testing transformer.
(e) To load charge split winding generators.
(f ) To inter-connect three supply systems operating at different
voltages.
Tertiary windings are mostly delta-connected.
Consequently, when faults occur on the primary or secondary
sides (particularly between lines and earth), considerable unbal-
ance of phase voltage, may be produced which is compensated by
large circulating currents.
In case of single line-to-ground fault, either on primary or second-
ary sides, the zero sequence current flowing through the delta-con-
nected tertiary winding is given as
Ir ¥ 100
I(Z0 Tert.) = (2.19)
2 Z H - L + Z0 L - T (or Z0 H - T)
(Z0 L – T to be taken if the fault is on LV i.e. sec-
ondary side and Z0 H – T, to be taken if fault is on
HV i.e., primary side).
where Ir = Rated current in tertiary winding at rated capac-
ity of main (primary/secondary) winding.
ZH – L = Positive sequence impedance (in percentage) be-
tween primary and secondary windings at rated
capacity of main windings.
Z0L – T = Zero sequence impedance (in percentage) be-
tween secondary and tertiary windings at rated
Z0H – T = Zero sequence impedance (in percentage) be-
tween primary and tertiary windings at rated
Thus the reactance of the winding must be such as to limit the
circulating current to that value, which can be carried by the cop-
per, otherwise the tertiary windings may overheat and mechani-
cally collapse under fault conditions.
2.14.2 Tertiary Windings in Star/Star Transformers

Star/star transformers comprising single-phase units, or three-phase
five-limb core-type units, suffer from the following disadvantages:
(a) that they cannot readily supply unbalanced loads between
line and neutral, and
(b) that their phase voltage may be distorted by third harmonic
emfs.
By use of delta-connected tertiary windings, induced currents are
caused to circulate in it, apportioning the load more evenly over the
three phases. The delta-connected tertiary provides a path for the
third harmonic currents.
2.14.3 Rating of Tertiary Winding

Rating of tertiary winding depends upon its use. If it has to supply
additional loads, its winding cross-section and design philosophy is
decided as per load and three-phase dead short-circuit on its termi-
nal with power flow from both sides of HV and MV.
In case it is to be provided for stabilization purposes only, its

cross-section and design has to be decided from the thermal and
mechanical considerations for the short duration fault currents
during various fault conditions single, line-to-ground fault being
the most onerous.
2.15 Parallel Operations of Transformers
When operating two or more transformers in parallel, their satisfac-

tory performance requires that they have
(a) The same voltage-ratio;
(b) The same per unit (or percentage) impedance;
(c) The same polarity;
(d) The same phase sequence and zero relative phase displace-
ment.
Of these, (c) and (d) are absolutely essential, and (a) must be
satisfied to a close degree. There is more latitude with (b), but the
more nearly it is true, the better will be the load division between
the several transformers.
(a) Voltage Ratio

If voltage readings on the open secondaries of various transformers,
to be run in parallel do not show identical values, there will be
circulating currents between the secondaries (and therefore
between primaries also) when the secondary terminals are connected
in parallel. The impedance of transformers is small, so that a small
percentage voltage difference may be sufficient to circulate a consid-
erable current and cause additional I2R-loss. When the secondaries
are loaded, the circulating current will tend to produce unequal
loading conditions and it may be impossible to take the combined
full-load output from the parallel connected group without one of
the transformers becoming excessively hot.
Case I: Equal Voltage Ratios
The equivalent circuit of two transformers running in parallel are
shown in Fig. 2.17. (Magnetizing current of both the transformers
have been assumed more or less equal, resulting in voltages
coincident in phases also.)
Let Z1, Z2 = the impedances of the two transformers
I1, I2 = their currents
Transformer 1
Z1
Z2
Transformer 2
Primary (a) Secondary
Z1 Z1
Z2 Z2
(b) (c)
Figure 2.17 Transformers in parallel.
V = common terminal voltage

I = total combined current
v = voltage drop across the impedances in parallel
From Fig. 2.17, v = I1Z1 = I2Z2 = IZ12, where Z12 is the impedance of
Z1 and Z2 in parallel.
Therefore,
Z1 Z2
Z12 =
( Z1 + Z2 )
So I1 =
v
=
IZ12
=I
LM
Z1 Z2 OP
Z1 Z1 Z1 ( Z1 + Z2 )
N Q
IZ2
=
( Z1 + Z2 )
Similarly,
I2 =
LM IZ OP 1
N (Z + Z ) Q
1 2
Multiplying both sides by common terminal voltage V

IZ2 ¥ V1
VI1 =
( Z1 + Z2 )
and VI2 = VI
LM Z OP 2
–3
N (Z + Z ) Q
1 2
Writing VI ¥ 10 = S, the combined load kVA, the kVA carried by
each transformer is
S1 = S
LM Z OP 2
N (Z + Z ) Q
1 2
and S2
L Z OP
= SM 1
(2.20)
N (Z + Z ) Q
1 2
These expressions are complex, so that S1 and S2 are obtained in
magnitude and phase angle.
Case II: Unequal Voltage Ratio
Let = E1, E2 = no-load secondary emf’s
and Z = load impedance at secondary terminals
Other notations are taken as in case I above.
Then v = IZ12 = (I1 + I2) Z12
Also the emf’s of the transformers will be equal to the total drops in
their respective circuits.
E1 = I1Z1 + (I1 + I2)Z
and E2 = I2Z2 +(I1 + I2)Z (2.21)
Whence
E1 – E2 = (I1Z1 – I2Z2)
as might be expected from this expression
I1 =
bE 1 - E2 + I2 Z2 g
Z1
giving I1 in terms of I2. Substituting this Eq. (2.21) above for E2 gives
E2 = I2Z2 +
LM b E 1 g
- E2 + I2 Z2 OP
+ I2 ¥ Z
MN Z1 PQ
Whence
I2 =
E2 Z1 - ( E1 - E2 ) Z U|
Z1 Z2 + Z( Z1 + Z2 ) ||
and by symmetry V| (2.22)
E1 Z2 - ( E1 - E2 ) Z
I1 =
Z1 Z2 + Z( Z1 + Z2 )
||
W
It is clear from Eq. (2.21) that, on no-load there will be a vector,

circulating current between the two transformers, of amount
I1 = – I2 = (E1 – E2)/(Z1 + Z2)
a result obtainable from Eq. (2.22) by writing
Z = • (infinity)
On short circuit the expected result from Eq. (2.22) is given by
I1 = E1/Z1
and I2 = E2/Z2
(b) Impedance
The currents carried by the two transformers are proportional to
their ratings if their numerical or ohmic impedances are inversely
proportional to those ratings and their per-unit impedances are
identical.
A difference in the quality factor (i.e. ratio of reactance to resis-
tance) of the per-unit impedance results in a divergence of the phase
angle of the two currents, so that one transformer will be working
with a higher, and the other with a lower power factor than that of
the combined output.
(c) Polarity
The primary and secondary windings of any individual transformer
may, under certain conditions of coil winding, internal connections
and connections to terminals, have the same or opposite polarity.
When, respective induced terminal voltages for primary and
secondary windings are in the same direction, the polarity of the
two windings is the same. This polarity is generally spoken as
subtractive. When, on the other hand, the induced terminal voltages
are in opposite direction, the windings are of opposite polarity
which is usually referred to as additive.
Figure 2.18 shows the test connections at (a) and (b) respectively
for single-phase transformers having subtractive and additive
polarity. When single-phase voltage is applied to terminals A1 and
a2, the measured voltage between terminals A1 and A2 shall be less
than applied voltage in case of (a) and greater than applied voltage
in case of (b).
For three-phase transformer the testing procedure is similar,
except that the windings must be excited from three-phase supply
and more voltage measurements have to be taken for determination
of exact polarity and phase sequence. Figure 2.18(c) shows the test
connections and results for a star/star-connected transformer with
subtractive polarity.
a2 a1
a2
A2 a1
Supply A2 A1
A1
(a) Subtractive polarity
a2 a1
A2 a1
a2
Supply A2 A1
A1
(b) Additive polarity
a2 A2
c2
b2
C2 C2
a2 b2 B2
Supply
A2 B2 C2
(c) Subtractive polarity
Figure 2.18 Test connection for determining polarity.
For parallel operation, the polarity should be the same. A wrong

polarity results in a dead short-circuit.
(d) Phase Sequence

The phase sequence, or the order in which the phases reach their
maximum positive voltages, must be identical for two paralleled
transformers. Otherwise, during the cycle, each pair of phases will
be short circuited.
Any mixture of internal connections can be worked out of it is
remembered that the primary and secondary coils on any one limb
have induced emf’s that are in time-phase. The several connections
produce various magnitudes and phases of secondary voltage; the
magnitude can be adjusted for parallel operation by suitable choice

of turn-ratio, but the phase divergence cannot be compensated.
Thus two sets of connections giving secondary voltages with a phase
displacement cannot be used for transformers intended for parallel
operation.
The following are typical of the connections for which, from the
view point of phase sequence and angular divergence, transformers
can be operated in parallel:
Transformer 1 Yy Yd Yd
Transformer 2 Dd Dy Yz
Internal connections of different windings can be done in a vari-
ety of ways, giving different vector groups. Various vector group ar-
rangements and their connection symbols are given in Indian Stan-
dard IS : 2026 (Part IV).
APPENDIX
Calculation of % Regulation and Efficiency
Transformer Specification 250 MVA, 15.75/240 kV, three-phase, 50 Hz

generator transformer having no-load loss 120 kW, load-loss 500 kW at
full-load at 75°C and reactance voltage drop 14% at rated load.
(a) Regulation at full-load, unity power-factor
R = 0.2%, X = 14%
cos q = 1.0, sin q = 0
Substitute values in Eq. (2.15)
Percentage regulation
(14 ¥ 1)2
= 0.2 ¥ 1.0 + = 1.18%
200
(b) Regulation at 3/4-load, 0.8 lagging power-factor
R = 0.15%, X = 10.5%
cos q = 0.8, sin q = 0.6
Substitute values in Eq. (2.15)
Percentage regulation
2
[(10.5 ¥ 0.8) - (0.15 ¥ 0.6)]
= 0.15 ¥ 0.8 + 10.5 ¥ 0.6 +
200
= 6.77%
(c) Efficiency at full-load, unity power-factor
Percentage efficiency
FG
= 1-
Total loss IJ
¥ 100
H Input power K
F
= 1-
500 + 120
¥ 100
I
GH 250 ¥ 103
JK
= 99.75%
(d) Efficiency at 3/4 load, 0.8 lagging power-factor
No-load loss = 120 kW
2
Load loss = 500 ¥ F 3I = 281.25 kW
H 4K
3
Input power = 250 ¥ 103 ¥ ¥ 0.8 = 150 ¥ 103 kW
4
Percentage efficiency
F
= 1-
281.25 + 120 I ¥ 100
GH 150 ¥ 103
JK
= 99.73%
REFERENCES
1. Say, M.G., The Performance and Design of AC Machines (Book), The
English Language Book Society and Sir Isaac Pitman & Sons,
London.
2. Puchstein, A.F., T.C., Lloyd and A.G., Conrad, Alternating Current
Machines, Asia Publishing House.
3. Hellman, Charles I, Elements of Radio (Book), D. Van Nostrand
Company, INC, New York.
4. Theraja, B.L., A Text Book of Electrical Technology, Sultan Chand &
Co., New Delhi.
C HAPTER 3
Materials Used
in Transformers
M.P. Singh
T.K. Ganguli
Apart from active materials like copper and cold rolled grain oriented
silicon steel, a number of ferrous, nonferrous and insulating materi-
als are employed for building up a transformer. Optimum utilization
of all materials in consonance with their electrical, mechanical,
physical, chemical and thermal characteristics is necessary for
obtaining a compact size transformer. Strict quality control measures
like testing of raw materials, therefore, become imperative. One
basic requirement for all materials used in an oil-filled transformer
is that they should be compatible with insulating oil and should not
react with or deteriorate oil.
Table 3.1 gives application, applicable national and international
standards of various materials used in a transformer.
3.1 Insulating Oil
Insulating oil forms a very significant part of the transformer

insulation system and has the important functions of acting as an
electrical insulation as well as a coolant to dissipate heat losses. The
basic raw material for the production of transformer oil is a low-
viscosity lube termed as transformer oil base stock (TOBS), which
is normally obtained by fractional distillation and subsequent
treatment of crude petroleum. Important characteristics of TOBS
given in Table 3.2 must be kept within permissible limits in order
to produce good insulating oils. TOBS is further refined by acid
treatment process to yield transformer oil.
Table 3.1 Materials Used in Power Transformer and Reactor
Sl. No Material Applicable Standards Application

and Grade
1 2 3 4
A. Insulating Materials
1. Transformer oil IS : 335, BS : 148 Liquid dielectric and coolant
IEC 296
2. Electrical grade paper
(i) Kraft insulating paper of medium IEC 60554–3–1 Layer winding insulation, condenser
air permeability core of oil impregnated bushing
(ii) Kraft insulating paper of high air IEC 60554–3–1 Covering over rectangular copper
permeability conductor
Covering over continuously trans-
posed copper conductor
Covering over stranded copper cable
Materials Used in Transformers
(iii) Crepe kraft paper BS 5626–3–3 Covering over flexible copper cable
Base paper as per (i) Insulation of winding lead
above Insulation over shield
(iv) Press paper IS : 8570, BS : 3255 Backing paper for axial cooling duct
(v) Kraft paper with aluminium Base paper as per (i) Line and common shield in winding
bands above
(vi) Crepe kraft paper with aluminium Base paper as per (i) Metallization of high voltage lead
foil above and shield
35
Contd.
Table 3.1 (Contd.)
36

and Grade
1 2 3 4
3. Pressboard
(i) Pressboard moulding from wet IEC 60641–3–1 Angle ring, cap, sector, snout, square
sheet or wet wood pulp tube, lead out and moulded piece of
intricate profile for insulating ends of
windings, insulation between windings
and numerous other applications
(ii) Soft calendered pressboard—solid Type C of Cylinder, barrier, wrap, spacer, angle
IS : 1576 washer, crimped washer, and yoke
IEC : 60641–3–1 insulation etc.
(iii) Soft pressboard—laminated BS EN 60763–1.2 Block, block washer, terminal-gear
Transformers, 2/e
cleat and support, spacer, etc.

(iv) Precompressed pressboard—solid IEC 60641–3.2 Dovetail block and strip, clack-band,
cylinder, warp, barrier, spacer, block,
block washer, corrugated sheet, yoke
bolt, washer, etc.
(v) Precompressed pressboard—laminated IEC 60763–3.1 Top and bottom coil clamping ring,
block, block washer, dovetail strip,
spacer, etc.
Contd.
Table 3.1 (Contd.)

and Grade
1 2 3 4
4. Wood and laminated wood

(i) Unimpregnated densified laminated IEC : 61061 Cleat and support, core/yoke clamp,
wood—low density wedge block, winding support block,
sector, core to coil packing, etc.
(ii) Unimpregnated densified laminated IEC : 61061 Coil clamping ring, block, cleat
wood—high density support, etc.
5. Insulated copper conductor and cable
(i) Paper covered rectangular copper IEC : 60317 For making different types of windings
conductor IS : 13730
(ii) Paper covered continuously trans- IEC : 60317 For making different types of windings
posed copper conductor
(iii) Paper covered stranded copper IS : 8572 conductor to For making lead and terminal
cable IS : 8130, IEC : 60228

(iv) Crepe paper covered flexible Conductor to For making lead and terminal req-
copper cable IS : 8130, IEC : 60228 uired to be bent to a small radius
(v) PVC insulated copper cable—single IS : 1554, BS : 6346 Control wiring in marshalling
and multicore IEC : 60502 box, nitrogen sealing system
Contd.
37
Table 3.1 (Contd.)
38

and Grade
1 2 3 4
6. Insulating Tape
(i) Cotton tape IS : 1923 For various taping purposes
(ii) Cotton newar tape — For taping and banding
(iii) Glass woven tape IS : 5352, IEC : 61067–1 Used in core bolt insulation
(iv) Woven terylene tape IS : 5351, IEC : 61068–1 For taping purposes at places req-
uiring higher strength
(v) Polyester resin impregnated — Banding of transformer cores
weftless glass tape
7. Phenolic laminated paper base sheet IS : 2036, BS : 2572 Terminal-gear support and cleat, gap
filler in reactor, tap changer components
Transformers, 2/e
8. Phenolic laminated cotton fabric sheet IS : 2036, BS : 2572 Terminal board, for making core duct,
support and cleat
B. Sealing Materials
9. Synthetic rubber bonded cork IS: 4253 (Part II) As gasket in different places to prevent
oil leakage from joints viz. tank rim,
turret opening, inspection cover and
with mounting flange of various
fittings, etc.
Contd.
Table 3.1 (Contd.)

and Grade
1 2 3 4
10. Nitrile rubber sheet and moulding BS : 2751 As gasket in different places to prevent
oil leakage from joints, ‘O’ ring in bush-
ings, moulded component in fittings
C. Ferrous Materials
11. Cold rolled gain oriented silicon steel BS : 6404/ASTM For making transformer core
(CRGO) A876M/ DIN 46400
12. Cold rolled carbon steel sheet IS : 513 For making radiator
ASTM A 620 M
BS : 1449–1.1
13. High tensile strength structural steel plate IS : 8500 Core clamp plate, anchoring and
clamping core to bottom tank
14. 1.5 % nickel-chromium-molybdeum IS : 5517 Lifting pin, roller shaft

steel bar and sections hardened and
tempered
15. Austenitic chromium nickel steel IS : 6911, BS : 1449 Turret opening, non-magnetic insert,
titanium stabilized plate (stainless steel) etc. to neutralize the effect of eddy
currents
16. Stainless steel sections (austenitic) IS : 6603, BS : 970 Non magnetic bar for high current
applications.
39
Contd.
Table 3.1 (Contd.)
40

and Grade
1 2 3 4
17. Structural steel—standard quality (plate, IS : 2062 Tank, end frame, clamp plate, ‘A’ frame
section, flat, bar, channel, angle, etc.) for radiator, conservator, turret, cable
box and for other structural purposes
18. Bright steel bar and sections—cold drawn IS : 7270 Threaded and machined components
D. Non-Ferrous Materials
19. High conductivity copper
(i) Sheet, strip, foil—hard and soft IS : 1897 U|
(ii) Rod IS : 613 || For various current carrying appli-
(iii) Tube BS : 1977 cations, e.g. bushing conductor,
Transformers, 2/e
(iv) Casting and forging BS : EN 1982

|V terminal lead, divertor and selector
(v) Tinned foil IS : 3331 || contacts of on-load tap changers,
(vi) Flexible cable IS : 8130, IEC : 60228 winding shield, cable box components,
(vii) Flat flexible braid — off-circuit switch items, etc.
||
20. Copper alloys W
(i) Free machining brass rod, square IS : 319 Tie rod and for making different
and hexagon components
(ii) Phosphor bronze rod IS : 7811 Tapchanger components
(iii) Nickel silver strip IS : 2283 For making winding shield
Contd.
Table 3.1 (Contd.)

and Grade
1 2 3 4
21. Aluminium
(i) Aluminium alloy plate Alloy 54300 M Flange in bushing, cable box, other non-
(NP 8—M) of IS : 736 magnetic applications
(ii) Aluminium plate (99.0 per cent) Alloy PIC of IS : 736 Shielding of reactor tank,
(iii) Aluminium foil — Condenser layer in bushings
41
Table 3.2 Characteristics of TOBS
Sl. No Characteristic Requirement

1. Viscosity at 40°C 9–14 cSt.
2. Pour, point, Max. – 9°C
3. Flash point (Pensky—Marten closed
cup method) Min. 145°C
3.1.1 Chemical Composition
Transformer oil consists of four major generic class of organic

compounds, namely, paraffins, naphthenes, aromatics and olefines.
All these are hydrocarbons and hence insulation oil is called a pure
hydrocarbon mineral oil. For good fresh insulating oil, it is desirable
to have more of saturated paraffins, less of aromatic and/or naphthenes
and none of olefines. However, for better stability of properties, it is
necessary to have optimum aromatic and/or naphthenic hydrocar-
bons. Such an optimum balance is struck by a carefully controlled
refining process. Depending upon the predominance, oil is usually
termed as of paraffinic base or naphthenic base.
3.1.2 Characteristics of Oil and their Significance
Table 3.3 gives characteristics of oil as per IS : 335. The typical

approximate values (all relating to 60°C) of some other physical
properties of oil considered in design calculations are given in Table
3.4. Figures in parenthesis indicate the approximate temperature
coefficient (per °C rise) for the property concerned. The significance of
various tests conducted on oil to evaluate its quality is given below.
A. Physical Properties
(i) Density. This test has special significance when transformer
is operated in a very low temperature zone. The maximum
value to density fixed at 29.5°C ensures that water in the
form of ice present in oil remains at the bottom and does not
tend to float on the oil up to a temperature of about –10°C.
(ii) Interfacial tension (IFT). This is a measure of the molecular
attractive force between oil and water molecules at their in-
terface. This test provides a means of detecting soluble polar
Materials Used in Transformers 43
Table 3.3 Schedule of Characteristics of Insulating Oil
Sl. No. Characteristic Requirement as per

IS : 335
1. Density at 29.5°C, max 0.89 g/cm3
2. Kinematic viscosity at 27°C, max. 27 cSt
3. Interfacial tension at 27°C min. 0.04 N/m
4. Flash point, Pensky—Marten 140°C
(closed), min.
5. Pour point, max. –6°C
6. Neutralization value (total acidity), 0.03 mg KOH/g
max.
7. Corrosive sulphur Non-corrosive
8. Electric strength (breakdown
voltage) min.
(a) As received 30 kV (rms)
(b) After filtration 60 kV (rms)
9. Dielectic dissipation factor 0.002
(tan-delta) at 90°C, max.
10. Specific resistance (resistivity), min
(a) At 90°C 30 ¥ 1012 W cm
(b) At 27°C 1500 ¥ 1012 W cm
11. Oxidation stability
(a) Neutralization value after 0.40 mg KOH/g
oxidation, max.
(b) Total sludge after oxidation, max 0.10% by weight
12. Presence of oxidation inhibitor The oil shall not contain
antioxidant additives
13. Water content (as received), max 50 ppm by weight
Table 3.4 Physical Constants of Oil
Sl. No. Property Value

1. Permittivity 2.2 (–0.001)
2. Specific heat 2.06 kJ/kg°C (0.0038)
3. Thermal conductivity 0.12 W/m°C
4. Coefficient of expansion 0.00078/°C
5. Mean density correction factor over the 0.00065/°C
normal range of operating temperature
contaminants and products of deterioration, which decrease

molecular attractive force between oil and water. It is con-
sidered that IFT gives an indication of degree of sludging of
oil (Fig. 3.1).
Sludge
formation zone
100
and found containing sludge
Percentage of oil examined
75
50
25
0 5 10 15 20 25 30 35
Interfacial tension, ¥ 10–3 NM
Figure 3.1 Presence of sludge in oils related to the interfacial

tension values at an ambient temperature of 27°C.
(iii) Moisture content. The amount of free and dissolved water

present in the oil is its moisture content and is expressed in
ppm (parts per million by weight, i.e. mg/kg). Presence of
moisture is harmful since it adversely affects the electrical
characteristics of oil and accelerates deterioration of insulat-
ing paper.
(iv) Flash point. It is the temperature at which oil gives so much
vapour that this vapour, when mixed with air, forms an ignit-
able mixture and gives a momentary flash on application of
flame under prescribed conditions. A minimum flash point is
specified in order to prevent the risk of fire that might result
by accidental ignition.
(v) Viscosity. It is a measure of oil resistance to continuous flow
without the effect of external forces. The oil must be mobile,
as heat transfer in transformers occurs mainly by convection

currents. Since viscosity increases with decrease in tempera-
ture, it is necessary that viscosity be as low as possible at low
temperatures.
(vi) Pour point. The temperature at which oil will just flow under
the prescribed conditions is known as the pour point. If the oil
becomes too viscous or solidifies, it will hinder the formation
of convection currents and thus cooling of equipment will be
severely affected.
B. Electrical Properties
(i) Electric strength (breakdown voltage). BDV is the voltage at
which breakdown occurs between two electrodes when oil is
subjected to an electric field under prescribed conditions.
Electric strength is the basic parameter for insulation system
design of a transformer. It serves to indicate the presence
of contaminating agents like moisture, fibrous materials,
carbon particles, precipitable sludge and sediment.
(ii) Resistivity (specific resistance). This is the most sensitive
property of oil requiring utmost care for its proper determi-
nation. Resistivity in W cm is numerically equivalent to
the resistance between opposite faces of a centimetre
cube of the liquid. Insulation resistance of windings of a
transformer is also dependent upon the resistivity of oil. A
low value indicates the presence of moisture and conductive
contaminants.
(iii) Dielectric dissipation factor (DDF). DDF is numerically
equal to sine of the loss angle (approximately equal to tangent
of loss angle for dielectrics) and is a good tool to indicate the
quality of an insulation. A high value of DDF is an indication
of the presence of contaminants or deterioration products
such as water, oxidation products, metal soaps, soluble
varnishes and resins.
C. Chemical Properties
(i) Neutralization value (total acidity). It is a measure of free
organic and inorganic acids present in the oil and is expressed
in terms of milligrams of KOH required to neutralize the
total free acids in one gram of oil. Oxidation of oil in service
is a consequence of reaction between hydrocarbons present
in the oil and oxygen. The oxygen may be atmospheric since

oil comes into contact with the atmospheric air during
breathing of transformer, or may have been dissolved in oil if
oil is not degassed properly, or may be liberated due to
the effect of heat on cellulose insulation. Oxidation of oil
is a chain reaction by which organic acids and sludge are
formed. Copper present in a large quantity in transformer,
acts as a strong catalyst in oxidation. Hence, as far as
possible, no bare copper is allowed to be used in power
transformers. Copper is covered with paper or enamel coated
or electrotinned. The products of oxidation are injurious to
the insulation system of transformer. Acids formed give rise
to formation of sludge which precipitates out and deposits
on windings and other parts of transformer. This causes
hindrance to proper oil circulation and heat dissipation. The
acids also encourage deterioration of cellulose insulation,
i.e. paper, pressboard and wood. Water is produced during
oxidation which reduces electric strength of oil and also
accelerates corrosion of metals and deterioration of insulating
materials. Hence the measurement of total acidity is
the most convenient and direct method of assessing the
capability of oil for nonformation of acids during service.
(ii) Oxidation stability. This test is the measure of neutralization
value and sludge after oil is aged by simulating the actual
service conditions of a transformer. The oxidation stability
test is very important for new oil but not for oil in service
and shows the presence of natural inhibitors which impart
anti-oxidation characteristics to oil.
(iii) Sediment and precipitable sludge. These are oil deterioration
products or contaminants which are insoluble after dilution
of the oil with n-heptane under prescribed conditions. How-
ever, precipitable sludge is soluble in the solvent mixture of
equal parts of toluene, acetone and alcohol but sediment is
insoluble in this solvent mixture. These contaminants are
determined for oils in service. Oil is considered unsatisfac-
tory for use if sediment or precipitable sludge is detected.
(iv) Corrosive sulphur. Crude petroleum usually contains sulphur
compounds, most of which are removed during the refining
processes. This test is designed to detect any traces of free
corrosive sulphur that may be present in oil. Presence of

corrosive sulphur in oil will result in pitting and black
deposit on the surface of bare copper used in transformer,
which will adversely affect the dissipation of heat and
consequently performance of the equipment.
3.1.3 Stability of Oil
Paraffinic base insulating oils available indigenously have shown

deterioration of electrical properties during use, at a rate which is a
little bit faster than of naphthenic base oils produced in the USA,
the UK and other countries. Efforts are afoot to produce paraffinic
base oil in the country having stability comparable to naphthenic
base oil.
3.2 Insulating Paper
Paper is a fabric made from vegetable fibres which are felted to form
a web or sheet. The fibrous raw materials are obtained from plants
including cotton, hemp, manila, straw and coniferous/deciduous
trees. The cell or fibres of such plants consist mainly of cellulose.
The molecular formula for cellulose is (C6H10O5)n. The degree of
polymerization of the molecular unit indicated by the letter n, varies
widely in various plants. The value of n can go up to 2500 or more for
cotton fibres and up to 1200 or more for wood pulp. Various other
materials, e.g. lignin, hemicellulose, mineral matter and resins are
associated with the cellulose in the fibres. These contaminants
are removed by sulphate process treatment of wood pulp and
careful water washing. The different types of paper used and their
application are given in Table 3.1. Terminal-gear and high voltage
leads are normally insulated using crepe kraft paper to obtain
higher flexibility and smaller radii at bends. These papers have
high stretchability varying between 50 to 100%.
3.2.1 Characteristics of Insulating Papers

and their Significance
Characteristics of different papers are given in Table 3.5 and their

Table 3.5 Schedule of characteristics of Electrical Grade Papers
Sl. Requirement
No. Characteristic Kraft paper Kraft paper Press
of medium air of high air paper
permeability permeability
1 2 3 4 5
I. Physical Properties
1. Substance, g/m2 100 ± 5 for 100 ± 5 for —
100 mm thk. 125 mm thk.
2. Density g/cm3 0.95 to 1.05 0.75 to 0.85 1.0 to 1.3
3. Moisture content, 8.0 8.0 8.0
% (max.)
4. Oil absorption, % — — 3 to 21%
(min) depending
upon density
5. Water absorption, 5 10 —
% (min.)
6. Air permeability 0.2–0.5 0.5–1.0 —
mm/Pa S
II. Mechanical Properties
7. Tensile strength
(expressed as tensile
index), N m/g (min.)
Machine direction 83 78 69
Cross direction 30 28 24
8. Elongation at break,
%, (min.)
Machine direction 2 2 —
Cross direction 4 4 —
9. Internal tear- Substance Tear resistance
ing resistance (g/m2) MD XD MD XD
(expressed as ≥40–80 5 6 5 6 —
tear index), > 80–120 6 7 6 7 —
mN m2/g, >120 8 9 — — —
(min.)
10. Heat stability 20 20 60
(decrease in bursting
strength), % (max.)
Contd.
Table 3.5 (Contd.)
Sl. Requirement
No. Characteristic Kraft paper Kraft paper Press
of medium air of high air paper
permeability permeability
1 2 3 4 5
III. Electrical Properties
11. Electric strength
(BDV), kV/mm, (min.)
(i) In air at 90°C 7.5 7.0 10
(ii) In oil at 90°C — — 65
12. Dissipation factor of 0.003 0.003 —
unimpregnated paper
at 105°C (max.)
13. Conducting paths should be free from any conducting path
IV. Chemical Properties
14. Mineral ash, %, (max.) 1.0 1.0 2.0
15. Conductivity of 5% 10 10 13.5
aqueous extract
mS/m, (max.)
16. pH value 6–8 6–8 5.0–8.5
17. Chloride content of 50 50 —
aqueous extract,
mg/kg, (max.)
18. Conductivity of 10 10 —
organic extract,
nS/m, (max.)
V. Methods of Test IEC 60554–3 IEC 60554–3 IS : 8570
IS : 9335 (Part II) IS : 9335 (Part II)
significance are described as follows:
A. Physical Properties
(i) Substance (grammage). It is the ratio of mass to the area and
is a fundamental parameter which influences most of the
mechanical and electric properties. This is a basic property
for paper making process.
(ii) Density. It is a function of thickness and grammage of
paper. Density is also a basic property for setting process
parameters for the manufacture of paper. Papers in density

range of 0.6 to 1.3 g/cm3 are employed in transformer
manufacture.
(iii) Moisture content. Cellulosic fibres are hygroscopic. Water has
the effect of plasticizing the cellulose fibres and of relaxing
and weakening the interfibre bonding. Hence moisture
content has a significant effect on many properties of paper.
(iv) Oil and water absorption. Paper attains a high value of
electric strength when impregnated in oil under vacuum,
since place of moisture liberated during vacuum drying is
taken up by oil in the molecular structure of the paper. Oil
absorption is dependent upon density and air permeability
of paper. Sometimes as a substitute to oil absorption, water
absorption is monitored.
(v) Air permeability. It is the measure of the rate at which paper
allows air to penetrate through it. It is influenced by both
the internal structure and the surface finish of the paper.
Electric strength of the paper is inversely proportional to its
air permeability.
B. Mechanical Properties
(i) Tensile strength and elongation. Paper should be able to
withstand tension exerted during its wrapping over layer
winding coils. Paper tape is wound over conductors and
cables under specified tension at a high speed by paper
lapping machines. If specified tensile strength is met,
breakages of paper will not occur in paper lapping operation.
It has been proved theoretically that the displacement of
windings during short circuit of a transformer depend very
much on the Young’s modulus of insulating materials used in
windings which shows the importance of tensile strength
and elongation tests.
(ii) Internal tearing resistance. This gives the load under which
paper will tear off under specified conditions and shows its
capability against tearing.
(iii) Bursting strength. This is the pressure required to burst a
disc of paper which is gripped firmly around its periphery,
and to one side of which pressure is applied at a uniform
rate, using liquid as a medium. This test also gives an idea
about the mechanical strength of paper.
(iv) Heat stability. It is the ability of paper to withstand thermal

stresses during service life of a transformer and is determined
by measuring decrease in internal tearing resistance, burst-
ing strength and degree of polymerization after subjecting it
to accelerated temperature.
C. Electrical Properties
(i) Electric strength. This is a basic parameter in deciding the
insulation system design of a transformer. Electric strength
depends on density and air permeability of paper.
(ii) Dissipation factor. It is a good tool to indicate the quality
of a dielectric. A high value of dissipation factor shows
presence of contaminants or depolymerization of paper.
(iii) Freedom from conducting paths. Electric strength of paper is
adversely affected by the presence of conducting paths and
hence these are undesirable impurities.
D. Chemical Properties
The conductivity of aqueous extract is a measure of the electrolytes
present in the paper which is extracted by gently boiling paper in
distilled water. The electrolytic impurities are present as ionizable
acids, bases, salts or a mixture of these. This property indicates
chemical purity of paper and is a basic parameter for paper making
process. Insulation resistance of paper is dependent on this
property. Conductivity of organic extract, pH value of aqueous ex-
tract, ash content and chloride content of aqueous extract are also
measured to check purity. Mineral ash, solid residue remaining af-
ter complete combustion of paper are injurious to paper and should be
within the permissible limits.
3.3 Pressboard
Pressboard is a widely used insulating material for making a variety

of components used in electrical, mechanical and thermal design of
transformers. Types of pressboard used and their applications are
given in Table 3.1. Like paper, pressboard is also made entirely
from vegetable fibres, whose cells contain mainly cellulose. Press-
board is manufactured using the following raw materials:
(a) Sulphate wood pulp

(b) Cotton
(c) Mixture of sulphate wood pulp and cotton or jute hemp
(d) Mixture of cotton and jute hemp
3.3.1 Characteristics of Pressboard
The characteristics of different qualities of pressboard commonly

used are given in Table 3.6. Solid pressboard up to 6 mm thick
calendered and up to 8 mm thick precompressed is made. Precom-
pressed pressboard is manufactured by pressing wet sheet at an
elevated temperature. For higher thickness, laminated calendered
and precom-pressed pressboard is manufactured by building up
laminae of solid pressboard using an adhesive (e.g. polyester resin,
polyvinyl alcohol, casien glue) for bonding under pressure and high
temperature.
3.3.2 Moulding
The most difficult practical insulation problems in high voltage

transformers occur at the ends of the windings and at the leadouts
from the windings. Pressboard moulded components can be made
to any required shape to follow the contour of the equipotentials
determined from the field plot. Angle rings and caps are the widely
used mouldings (Fig. 3.2 and Fig. 3.3). Other types of mouldings used
in EHV transformers are snouts, angle sectors and leadouts, etc.
Evaluation of moulding is another difficult task due to intricate
profiles. These cannot be tested without destroying them. Normally,
bend/neck of the component is the weakest point. Hence, these are
tested on regular, irregular and neck portions to ensure the overall
quality.
3.3.3 Advantage of Pressboard Over Other

Solid Insulating Materials
In EHV transformers care is taken to use such insulating materials

in stress zones which will have the least partial discharge. Synthetic
resin bonded paper based laminates and cylinder and laminated
wood are prone to give more P.D. in such zones compared to
pressboard, since air voids are trapped due to extensive use of resin
Table 3.6 Schedule of Characteristics of Different Types of Pressboard
Requirement
Sl. Precompressed Soft Precompressed Soft
No. Characteristic solid calendered laminated calendered
solid laminated
1 2 3 4 5 6
Composition 100% wood pulp Cotton, jute 100% wood pulp Cotton, jute
and wood pulp and wood pulp
1. Density, g/cc 1.1–1.3 0.9–1.1 1.2–1.3 0.95–1.15
2. Moisture content, % (max.) 6 8 5 7.5
3. Oil absorption, % (min.) 9 20 5 12
4. Shrinkage, % (max.)
(i) In Air
Machine direction 0.5 1.0 0.4 1.0
Cross direction 0.7 1.5 0.6 1.5

Perpendicular 5.0 6.0 4.0 5.0
(ii) In Oil
Machine direction — 0.3 — 0.5
Cross direction — 0.3 — 0.5
Perpendicular — 1.5 — 0.5
5. Cohesion between plies The specimen shall not readily split by delamination, i.e. between two adjacent
plies and the exposed torn surfaces shall have a distinctly hairy or ragged
53
appearance.
Contd.
Table 3.6 (Contd.)
54
Requirement
solid laminated
1 2 3 4 5 6
6. Tensile strength, MPa (min.)
(i) Machine direction 100 47 — —
(ii) Cross direction 75 22 — —
7. Cross breaking strength, MPa (min.)
Transformers, 2/e
(i) Machine direction NA NA 100 32.5

(ii) Cross direction NA NA 85 30
8. Compressibility, % (max.)
(i) In air 4–10 13 3 9
(dependent on thickness)
(ii) In oil — 20 — 13
9. Electric strength in oil 90°C kV/mm (min.)
(i) 1.5 mm thick 35 BDV 15.3 for one min. U|
(ii) 3 mm thick 30 BDV 11.3 for one min. 70 kV for one min. on
(iii) 6 mm thick 30 BDV 8.3 for one min. 25 mm width (edgewise)
V|
(iv) 8 mm thick 25 BDV — W
Contd.
Table 3.6 (Contd.)
Requirement
solid laminated
1 2 3 4 5 6
10. Conductivity of 5% aqueous 10 11 10 15
extract, mS/m, (max.)
11. pH value of aqueous extract 5.5–9 7–9.5 5–8 5–9
12. Mineral ash, % (max.) 1 2 2 2
13. Effect on insulating oil
(i) Increase in acidity, mg N.A. N.A. 0.1 0.1
KOHg, (max.)
(ii) Increase in sludge content, N.A. N.A. 0.05 0.05
% (max.)
V. Method of Test IEC 60641–3–1 IS : 1576 IEC : 60763-3.1 BS : 5354
N.A. = Not Applicable.
55
Open scarf
Figure 3.2 Typical pressboard angle rings.
as the bonding material. Oil absorption in resin bonded laminates

and cylinders is negligible and hence these are electrically inferior
to pressboard. Efforts, are made to use cylinders, winding cleats and
supports, etc., made of pressboard as far as possible.
3.3.4 Significance of Pressboard Characteristics
Properties like density, tensile strength, elongation, ash content, pH

of aqueous extract, conductivity of aqueous extract, oil absorption,
Open scarf
Figure 3.3 Typical pressboard caps.
moisture content, freedom from conducting particles and electric

strength, described in Sec. 3.2 are also determined on pressboard.
Apart from the above, the details of additional tests conducted for
the evaluation of pressboard are given below:
(i) Compressibility. It is a very important property and is a
measure of reduction in thickness of dried specimen when
compressed under load, under prescribed conditions. Trans-
former windings are designed based on expected compressed
thickness of dovetail blocks used for providing radial ducts
between turns and sections. If compressibility of pressboard
is more than the value taken in the design, the coil height
will become less than the design value after vacuum drying
and oil impregnation of transformer and this reduction can
lead to serious problems of short circuit forces if not rectified.
(ii) Shrinkage. It is a measure of reduction in length, width
and thickness of specimen after it is dried under specified
conditions. Effect of shrinkage is also taken into consideration
in the design.
(iii) Cross-breaking strength. This is determined to check the
proper adhesion of laminae in a laminated pressboard and
ensures against delamination.
(iv) Electric strength edgewise. It measures the creep strength
across the glue line of laminated pressboard.
3.4 Wood
Wood based laminates are manufactured from selected veneers

(mostly 2 mm thick) obtained from various timbers. The veneers
are dried and partially or fully impregnated with neutral phenol
formaldehyde resin depending upon type of laminate required and
then densified under heat and pressure. The placement of veneers
one over the other is such as to obtain the desired grain orientation
to achieve specific mechanical strength in the required direction. The
following types of wood based laminates are used in transformers
whose applications are given in Table 3.1 and properties in Table 3.7.
Type Direction of Veneers
(a) Unimpregnated densified Alternate veneers with grain
laminated wood—low density directions at right angles to
and high density. each other.
(b) Resin impregnated densified All veneers have grain in the
laminated wood type V. same direction.
(c) Resin impregnated densified Same as of (a).
laminated wood type VI.
The areas which require higher mechanical and lower electric
strength, densified laminated wood is used for making a variety of
insulation components like coil clamping ring, cleat, support, core
and yoke clamp, wedge block, bolt and nut, terminal board and core
step filler, etc. Haldu and teak wood seasoned planks are also used
as a filler material between core limbs and enveloping coils and as
yoke-step fillers.
3.4.1 Evaluation of Wood
Details and significance of tests conducted on densified laminated

wood as per Table 3.7 have been described is Secs. 3.2 and 3.3. The
quality of laminated wood may vary widely from batch to batch and
sometimes even sheet to sheet. The reason has been attributed to
many parameters, e.g. species of wood selected for veneers, their
location and age, number of joints in veneers to make up the required
width, gap in joints and defects in timber, etc. Hence, utmost care is
taken in their use.
3.5 Insulated Copper Conductor for Windings
The following types of conductors are used for making spiral, heli-
cal, continuous disc, layer and interleaved disc windings of trans-
formers described in Chapter 5.
(a) Paper covered rectangular copper conductor.
(b) Twin paper covered rectangular copper conductor bunched
together.
(c) Paper covered continuously transposed copper conductor
(CTC).
(d) Twin transposed copper conductor bunched together.
(e) Twin rectangular copper conductors bunched together and
provided with a common paper covering and a glued (epoxy
coated) paper strip between the two conductors.
(f) Epoxy coated continuously transposed conductor.
Conductors at (b) and (d) above are used to improve the winding
space factor. Individual conductor is covered with only three or four
layers of paper to electrically separate them instead of giving full
insulation as shown in Fig. 3.4. The two conductors/cables are kept
either in axial height or in radial depth of coils, thereby reducing
total coil axial height and radial depth respectively.
Conductors at (e) and (f) further improve the mechanical
strength of the winding, besides improving the space factor of the
Table 3.7 Characteristics of Densified Laminated Wood
Sl. Unimpregnated Unimpregnated

No. Characteristic laminated wood laminated wood
—low density —high density
1. Density, g/cm3 0.90 to 1.09 1.10 to 1.20
2. Moisture content, % (max.) 7 7
3. Oil absorption, % (min.) 9 6
4. Cohesion between laminae The specimen shall not readily split
by delamination and the exposed
torn surfaces shall have a distinctly
hairy or ragged appearance.
5. Tensile strength, MPa, (min.) 70 95
6. Cross breaking strength, MPa, (min.)
(i) Along the grain 90 112
(ii) Across the grain 67 85
7. Compressive strength, MPa, (min.)
Perpendicular to laminae 160 185
8. Shear strength, MPa, (min.) 35 50
perpendicular to laminae
9. Electric strength (one minute
proof) in oil at 90°C
(i) Flatwise, kV/mm (min.) 4 4
(ii) Edgewise, kV (min.) 60 60
10. Effect of wood on oil
(i) Increase in acidity, 0.1 0.1
mg KOH/g, (max.)
(ii) Increase in sludge content, 0.05 0.05
% (max.)
V. Methods of Test IS : 1998 IS : 1998
winding. These conductors are used to obtain better mechanical

withstand properties without sacrificing the electrical properties
which are difficult to meet in case single conductors are used.
Bare strip
Paper covering on individual strips
Paper covering on bunch

(a)
Paper covering on individual cable
Paper covering on bunched cable

Bare transposed cable

(b)
61
Figure 3.4 (a) Twin rectangular bunched copper conductor. (b) Twin transposed copper conductor.
However, due care is required to be taken for manufacturing

of windings with these conductors, as these conductors especially
conductors at (e) have limited shelf life and the glued (epoxy) strip is
susceptible to loose its mechanical properties at an exponential
rate when exposed to higher temperatures. The shelf life of these
conductors is expected to reduce by half for every 10°C increase in
temperature.
The bare strips are made from electrolytic tough pitch (ETP)
grade copper wire bars to IS : 191 with high conductivity (99.14%
IACS or 57.5 Sm/mm2) and annealed complying with requirements
of IS : 6160. The manufacturing operations involved are hot rolling,
pickling, shaving, drawing, flattening, annealing and finally covering
with requisite number of kraft paper layers. Specified radius is given
on the four corners of a strip to avoid sharp edges. Since transformer
windings are to withstand power frequency, impulse and switching
surge voltages during factory tests and operation, it is required that
a strip is perfectly smooth. The surface should be free from various
defects like spills, cracks, slivers, scratches, pits, black spots and
copper dust. Micro projection on the strip surface may puncture a
few layers of insulating paper. Surface defects are likely to produce
partial discharges due to nonuniformity of field, which may
ultimately lead to conductor insulation breakdown. This is the
reason for giving importance to electric strength test on these
conductors. Table 3.8 gives test voltage for electric strength test on
rectangular and transposed conductors as well as on paper covered
stranded and flexible cables.
High air permeability kraft paper with properties in accordance
with Table 3.5 is used as covering. All layers except the outermost
are butt wound, the outermost layer being overlap wound. Maximum
thickness of paper for butt wound layers is normally 75 mm. Increase
in dimension due to paper covering is measured by difference
method as well as by adding up the thickness of paper used in each
layer to ensure reasonable tightness of covering.
3.5.1 Continuously Transposed Copper Conductor (CTC)
CTC comprises an odd number of high conductivity annealed copper

strips insulated with a polyvinyl-acetal based enamel, arranged in
two side-by-side stacks with the individual enamelled conductors
Table 3.8 Test Voltage for Electric Strength Test on Rectangular/

Transposed Conductor and Flexible/Stranded Cable
Radial insulation Test voltage breakdown voltage

thickness (mm) in oil at 90°C
(kV rms)
0.5 20
1.0 35
1.5 48
2.0 61
2.5 72
3.0 83
3.5 93
4.0 104
4.5 113
5.0 123
7.0 160
10.0 199
continuously transposed throughout their length. Additional

insulation is provided by covering the transposed stack by a number
of paper layers, as for rectangular single conductors. Figures 3.5(a)
and (b) show the general view and constructional details and
Fig. 3.6 depicts the transposition of individual strips in a CTC.
Pitch of transposition is the distance between adjacent crossovers
measured on one side of the conductor and is normally 15 times the
individual strip width. At least one full transposition should be
made in a full turn of the winding. Hence, the minimum winding
diameter is decided by the relationship.
pitch of transposition ¥ no. of conductors
Dmin =
p
CTC offers the following technical advantages over single
rectangular conductor (economic aspects are also taken into account
for optimum design):
(a) Winding time is reduced, since the need for hand transposi-
tions is partially or completely eliminated.
(b) The winding space factor improves noticeably compared
with single rectangular conductors.
W
64
H Radial height of CTC.
W Axial width of CTC.
w t Thickness of individual strip (nominal)
w Width of individual strip (nominal)
tp te Thickness of enamel
2 tp Total thickness of paper added.
Zte
H t
Transformers, 2/e
(a) (b)
Figure 3.5 (a) Constructional details of continuously transposed conductor.

(b) General view of transposed-strip conductor (27 strips in parallel).
1 7 6
7 2 6 1 5 7
6 3 5 2 4 1
5 4 4 3 3 2
Figure 3.6 Method of transposition of strips in a CTC.

65
(c) By employing a number of small conductors which are

continuously transposed, eddy losses, due to circulating
currents between adjacent turns are reduced and hence
total copper-losses of transformer due to windings are also
reduced.
(d) Transposed conductor offers an improvement in the
mechanical strength of the winding.
3.6 Crepe Paper Covered Flexible Copper Cable
Tapping leads and high voltage leads are required to be given sharp
bends up to a radius of four to five times the overall diameter of
cable. The cables used have high flexibility due to use of a fairly
large number of thin wires. Clubbed with the benefits of crepe paper
mentioned in Sec. 3.2, these cables provide extra-high flexibility
and damage to paper insulation at sharp bends is eliminated.
3.7 Sealing Materials
For all sealing purposes, to avoid oil leakage at joints, gasket and
moulded components made from nitrile rubber or nitrile rubber
bonded cork are used. Characteristics of these materials are given
in Table 3.9. PTFE and polyamide washer, tape, etc. are used for
plugs and valves.
3.7.1 Nitrile Rubber Bonded Cork
This is the most widely used general purpose gasket. The cork
gasket is adaptable even to rough unmachined surface by virtue of
its compressibility under light seating loads. This gasket is made by
the compounding of the granular cork with medium nitrile butadiene
rubber. The cork used is clean, soft-grade type, uniformly granulated
and free from hard board, wood flour, dust and other foreign
materials. Nitrile rubber contributes to the ‘sealability’ factor that
enables the bonded cork to effect a tight seal under very light seating
loads and also permits considerable distortion of gasket without
crack and damage. The presence of cork allows compression without
excessive spreading associated with solid rubbers.
Table 3.9 Schedule of Characteristics of Gasket Materials
Sl. Requirement
No. Characteristic Nitrile rubber Nitrile
bonded cork rubber
I. Mechanical Properties
1. Hardness, IRHD 70 ± 10 60 ± 5
2. Tensile strength, 1.75 8.4
MPa (min.)
3. Elongation at break — 400
% (min.)
4. Compressibility at 25–35 —
2.8 MPa, % (max.)
5. Recovery % (min.) 80 —
6. Compression set, % (max.) 85 30
7. Flexibility No breakage —
through the
granules of cork
or separation of
granules should
occur.
II. Effect of Ageing in Air
8. Increase in hardness, N.A. 10
IRHD (max.)
III. Compatibility with Insulating Oil
9. Increase in acidity, 0.1 —
mg KOH/g (max.)
10. Increase in sludge 0.05 —
content, % (max.)
IV. Methods of Test IS : 4253 IS : 3400
3.7.2 Nitrile Rubber
This is used as gasket normally with a metal limiter to avoid exces-

sive compression and as sealing rings and mouldings, etc. Nitrile
rubber is a combination of butadiene and acrylonitrile rubber con-
forming to compound BA 60 of BS 2751. The resilience of nitrile
rubber makes it possible to reuse the gasket after opening a

gasketted joint. However, this requires machined flange surfaces.
For smaller flanged openings, a groove is cut into the flange for
seating of nitrile rubber gasket. In such a case, limiter is normally
not required.
3.8 Cold Rolled Grain-Oriented Electrical Steel Sheet (CRGO)
CRGO electrical steel with an approximate silicon content of 3% is

used for magnetic circuits of a transformer. The following features
influence selection of the type of steel sheet.
(a) Maximum magnetic induction to obtain a high induction
amplitude in an alternating field.
(b) Minimum specific core-loss for low no-load loss.
(c) Low apparent power input for low no-load current.
(d) Low magnetostriction for low noise level.
(e) High grade surface insulation
(f) Good mechanical processing properties.
CRGO made from a ferrous base present maximum magnetizability,
i.e. permitting a high induction. Iron crystallizes into a body-
centered cubic lattice, with the cube edges of the lattice pointing in
the direction of easiest magnetizability and lowest, core-loss. Grain
oriented electrical sheet consists of a silicon-iron alloy, with the
crystallites being predominantly oriented by means of a specific
manufacturing process, in such a way as to have four cube edges
pointing in the rolling direction and the diagonal plane being
parallel to the sheet surface. In this way the rolling direction becomes
the direction of maximum magnetic properties and approaching the
ideal properties of the individual crystallite. The more pronounced
the texture, the nearer the properties become to those of an
individual crystallite.
3.8.1 Grades of CRGO
CRGO steel for transformers is available in different sheet

thicknesses, ranging from 0.18 thickness, to 0.35 thickness, and
accordingly the maximum watt loss/kg of the material differs.

Further, based on the grain orientation they are also categorized as
conventional grain-oriented (CGO) and HI-B steels. HI-B steels have
low watt loss/kg when compared to CGO steel. The schedule of char-
acteristics of some typical grades commonly used in transformers is
given in Table 3.10.
Table 3.10 Schedule of Characteristics of CRGO Electrical Steel (Typical)
Sl. No. Characteristic Grade CGO Grade HI–B

1 2 3 4
I. Physical Checks
1. Thickness, mm (0.18, 0.23, 0.27, 0.30, 0.35)
2. Edge camber, mm (max.) 2 for 2000 mm sample length
3. Edge burr, mm (max.) 0.05 0.05
4. Waviness, (max.) 5 mm for 2000 mm long sample,
height of any wave not to exceed 1/80
of the length of a wave, or 1 mm,
which-ever is the greater.
5. Stacking factor, % (min.) 93.5 to 96.5 depending on thickness
II. Magnetic Properties
6. Specific core-loss, W/kg (max.)
at 50 Hz.
(i) At 1.5 T 0.69–1.11 —
(ii) At 1.7 T 1.17 to 1.57 0.95–1.17
7. Ageing test (increase in core- 3 3
loss after ageing at 150°C
for 14 days), % (max.)
8. Magnetic induction at mag- 1.75 1.85
netizing force of 800 A/m
(B8), T (min.)
9. Specific apparent power at 1.25 1.0
induction of 1.5 T, VA/kg
(max.)
10. Saturation induction, T 2.03 2.03
Contd.
Table 3.10 (Contd.)
Sl. No. Characteristic Grade CGO Grade HI–B

1 2 3 4

11. Resistance of surface 2 for 80% of the readings
coating, W (min.) 5 for 50% of the readings
IV. Mechanical Properties (Typical)
12. Ductility (Bend test) Min. one 180° bend without fracture
13. Tensile strength, MPa
(i) Longitudinal 350 330
(ii) Transverse 420 390
14. Yield point, MPa
15. Elongation, %
16. Hardness, HV (load 50 N) 175 170
V. Physical Properties (Typical)
17. Density, g/cm3 7.65 7.65
18. Specific electric resistance, 48 45
mW cm
19. Thermal conductivity at 25°C, 26 28
J/m sK
VI. Methods of Test BS 6404—Part 5
VII. Applicable standards BS 6404—Part 2
DIN 46400
JIS C 2553
ASTM 876 M
Specific core-loss is made up of hysteresis and eddy curren losses.

Hysteresis loss is usually reduced by making an improvement in
grain orientation. As the (100) pole figures in Fig. 3.7 show, the
mean deviation angle of the (100) axis from the rolling direction is
about 3° for HI-B, which is far lower than the average 7° for CGO.
Thus in terms of hysteresis loss, HI-B is one or two grades superior
RD RD
HI-B CGO
Figure 3.7 (100) pole figures showing the grain orientation of

HI-B and CGO.
Hysteresis loss Wh 15/50 and Wh 17/50 (W/kg)

0.1 0.2 0.3 0.4 0.5
1.70
0
/5
15
h
W
1.75
D Wh 15/50 0.12 W/kg
/50
1.80 17
Wh
Induction B8 (T)
1.85
1.90
1.95
D Wh 17/50 0.16 W/kg
2.00
0.05 0.10 0.15 0.20 0.25
Hysteresis loss Wh 15/50 and Wh 17/50 (W/lb)
Figure 3.8 Effects of B8 (induction at magnetizing force of 800

A/m) value on hysteresis loss.
Phosp
hate c
oating
Glass
film
n
)
io
00
a
ct
(1
re
)
00
di
a
g
(1
llin
a
Ro
)
00
0) a
(10
(1
Figure 3.9 Grain arrangement of GO—Heavy arrows show the

(100) direction.
than conventional material (Fig. 3.8). Eddy current loss depends to

a large extent on the sheet thickness, the frequency of the alternat-
ing field and to a minor degree on the electrical conductivity. Eddy
current loss is usually improved by applying a surface coating which
generally consists of a glass film and a phosphate coating (Fig. 3.9).
This surface insulation, uniformly coated on both sides withstands
stress relief annealing, without deterioration in its adhesion or
electrical insulation value, at temperatures up to 840°C. The mean
value of the insulation resistance of the coating is at least 10 W cm2
on each side.
Figures 3.10 and 3.11 show the improvements in core-loss of HI-B
steel over CGO when the induction is over 1.5 T. Therefore, HI-B
permits the manufacture of transformers with the same no-load
loss as before, but a higher induction amplitude with the resulting
economic advantages such as smaller size, less weight and reduced
amount of material as shown in Fig. 3.12.
Reduction in core-loss in CRGO and HI-B grade steel has been
possible with reduced thickness, and by change/modifications in
the metallurgical process. Now, with present day techniques, it is
further possible to reduce the core-losses by refining the magnetic
domain structure of the steel. This technology involves in further
subdivision of the width of the magnetic domains by physical
means, or by electrolytic etching, or by laser irradiation.
M65
M6
1.4
M55
1.3 M4H
CGO M5
1958 M45
Core-loss (w/kg)
1971 M3H
M4
1.2 M2H
M3H
1.1 HI-B
1971 M2H
M2H
1972 M1H
M1H
1.0
1975 M0H
0.8
0.28 0.30 0.35
Thickness (mm)
Figure 3.10 Improvement in core-loss.
Table 3.11 lays down the possible methods of reducing core loss:
However, when domain refined steel is used for manufacturing of
the transformer core, it is very important to keep in mind that this
material should not be subjected to annealing, unlike CGO and
HI-B grade steels, as this process would totally disturb the domain
refinement achieved, and would lose the purpose of its use, i.e. low
core-loss.
74
2.0
mm
0.30 mm
HI-B 0.30
1.8
CGO
1.6
1.4
Induction B (T )
1.2
Transformers, 2/e
HI-B 0.30 mm
CGO 0.30 mm
1.0
0.8
0.5 1.0 1.5 2.0
Core-loss (W/kg)
Figure 3.11 Comparison of B.W. curves between HI-B and CGO.

B = 1.4 T
B = 1.85 T
Figure 3.12 Schematic comparison between a three-phase trans-

former with a mean induction amplitude and one
with a high induction amplitude.
Table 3.11 Methods of Reducing Core-loss
Core-loss Core-loss reducing method

Reduction in hysteresis loss 1. Improvement in orientation.
2. Reduction in inclusions and
impurity.
3. Reduction in internal strain.
Reduction in Classical eddy 1. Reduction in thickness.
Eddy current current losses 2. Increase in silicon content to
increase resistivity.
Anomalous eddy 1. Grain size.
current-loss 2. Refinement of magnetic domain.
Alternatively, transformers can be manufactured with 5 to 20%

reduction in no-load loss depending upon the grade and induction
amplitude using HI-B steel.
3.8.2 Core-loss in Cross-grain Direction
If magnetization is applied in directions other than the rolling

direction, the core-loss increases substantially. Core-loss is more
than three-folds at 90° to the rolling direction and more than
four-folds at 60°. This aspect is taken care of while designing the
magnetic circuit of transformers. Core-loss of HI-B steel in 90°
direction is inferior to those of conventional steel. The B-H curves
in Fig. 3.13 indicate that HI-B has a high permeability. However,
saturation induction of both HI-B and CGO is comparable.
Due to lower specific apparent power of HI-B steel, 10 to 50%
reduction in transformer exciting volt-amperage is obtained.
3.8.3 Magnetostriction
For environmental protection, the noise level produced by the trans-

formers has recently been a significant factor for the user. Such
transformer noise is generated by the iron core of the transformer,
as well as from that made by the fans. In an ideal core, the lamina-
tions of which are pressed together in such a way as to prevent any
fluttering or shifting, the noise is generated by the magnetostriction
of the laminations in the alternating field, causing the core to vibrate
and thus act as an acoustic source. If the above mechanical
conditions of the core manufacture are not observed, other causes
(such as magnetic vibrations) may also be the reason for noise.
Magnetostriction, i.e. the change in configuration of a magnetiz-
able body in a magnetic field, leads to periodic changes in the length
of the body in an alternating magnetic field. The frequency of this
magnetostriction is twice as large as that of the existing alternating
field, i.e. when magnetizing an electrical sheet with 50 Hz the
fundamental frequency of magnetostriction is 100 Hz. The process
of change in length as a function of induction is not strictly a linear
one, so that the magnetostriction of an alternating field gives both a
fundamental and harmonic frequencies. The fundamental frequency
is predominant in an alternating magnetic field up to an induction
2.0
HI-B Assumed
1.8
CGO
1.6
Induction B (T )
1.4
1.2 HI-B 0.30 mm

CGO 0.30 mm
1.0
0.8
10 50 100 500 1,000 5,000 10,000
Magnetizing force H(A/M)
Figure 3.13 Comparison of D-C B-H curves between HI-B and CGO.
77
1.7 T, while as a result of the sophisticated process of alternating

contractions and expansions, the harmonic frequency prevails for
induction above 1.7 T. Magnetostricton is minimum in the rolling
direction, whilst in the 90°-direction it is maximum. Except for the
90°-direction, all other directions that present major deviations
from the rolling direction have a high percentage of harmonic
frequencies in the spectrum, which are particularly undesirable
because they generate harsh noise in frequency ranges, where the
human ear is especially sensitive. Therefore, for sheets used
in transformers, all transverse and oblique effects are kept to a
minimum and very high induction amplitudes are prevented in the
rolling direction.
Magnetostriction of HI-B is lower than that of CGO (Fig. 3.14)
and consequently 2 to 7 dB reduction in the noise level of trans-
former is achieved.
3.8.4 Stress Sensitivity
Mechanical stresses exert a significant effect on the magnetic

properties of CRGO electrical sheet. Such stresses are introduced in
a sheet by way of external forces and by way of plastic deformation
(internal stresses). External stresses are developed due to the
following:
(a) When the sheets of a core are exposed to tensile or
compressive stresses exerted by the fixings with the
longitudinal forces in the limbs and the yokes.
(b) Bending stresses, where the sheets are forced into waves (for
instance, as a result of heavily tightened clamp plates or end
frames).
(c) Bending stresses due to deflection (for instance, in the case
of non-uniform corner jointing).
(d) When use is made of wavy or bent sheets, which are
necessarily flattened in the core.
(e) When sheets cannot adopt their natural condition as a result
of friction existing between them.
Internal stresses occur along the cut edges during each cutting
operation and as a result of bending the sheet or subjecting it
to tension beyond the yield point (in slitting lines, continuous
furnaces, etc.). Bending stresses can be divided into components of
X10–8
2 HI-B 0.30 mm
CGO 0.30 mm
Magnetostriction DL/L
60 Hz
1
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Induction B (T )
Figure 3.14 Peak-to-peak magnetostriction against induction for HI-B and CGO.
79
tension and compression. Similarly in simple cases, internal

stresses can be resolved into pairs of tensile and compressive com-
ponents. A few examples are given in Fig. 3.15.
Tensile and compressive components

Bending in case of external mechanical stress
Bending Stress Stresses
Stress Forced flatt- between sheets
Forced ening of a Compression
waviness curved
Tension
steel
Com-
Tension pres-
A
True length
sion
Zones of internal stress
Tension Tension
B
Compr-
ession Tensile and compressive component
Tension
Compression in case of internal stress in a strip

section
True shape Frictional layer

Compression
Figure 3.15 Examples of mechanical stress in a sheet.
Pz DPz Relative change

DPs
Relative change in apparent power

absorption Pz and specific core-loss Ps
as a result of mechanical stress ‘0’
Magnetization and stress in rolling

direction
Ps
Pz
0
Ps
–10 –8 –6 –4 –2 0 2 4 6 8 10 MPa
Compression Tension
Figure 3.16 Core-loss and apparent power absorption of CRGOS

as a function of external mechanical stress.
Figure 3.16 Displays the specific core-loss Ps and the apparent

power Pz for an induction of 1.7 T and various levels of mechanical
stresses. Both Ps and Pz are highly dependent on the compressive
stress but only slightly dependent on the tensile stress. Thus, the
apparent power is more sensitive than the core-loss.
Figure 3.17 shows the dependence of magnetostriction upon the
tensile and compressive stresses for an induction of 1.7 T. Here too,
effect of compressive stress is much bigger than that of tensile
stress. Hence compression beyond a certain level occurring in
the plane of the sheet is harmful and therefore, stresses must be
removed by stress relief annealing. Sheets affected by waviness or
coil-set are maintained to a flat shape, since they are subject to elastic
stresses when being nested into the final shape of the transformer
cores due to pressure. Buckled sheets must be discarded. Stress
relief annealing of laminations is done in a continuous rolled
hearth-type furnace at 820 ± 20°C for approximately 2 minutes in
an air atmosphere and cooled quickly to ambient temperature.
Internal residual stress inherent in HI-B strip in as-delivered
condition is lower than that in CGO, thereby stress-relief annealing
may be omitted in many cases.
3.9 Structural Steel
Structural steel mainly in sheet and plate form is used in the

fabrication of transformer tank, radiator, conservator, clamp plate,
end frame, marshalling box, cable box, roller, turret and inspection
cover, etc. Normally, standard quality mild steel is conventionally
used for core clamp plates. However, it is economical to use high
tensile strength (HTS) steel plate, since by reducing the thickness
of clamp plates, magnetic core area can be increased, resulting in
improvement in the utilization factor. In such a case, lifting pins
required to connect clamp plate to yoke end frame are manufactured
from HTS bar (1.5% nickel-chromium-molybdenum) having a very
high tensile strength in the range of 900 to 1050 MPa.
Stainless steel (austenitic chromium nickel) plates are used as
cover for turrets of high current bushings to neutralize the effect of
eddy current. For medium currents, magnetic circuit perpendicular
to a current carrying conductor is broken by welding stainless steel
inserts. These plates are titanium stabilized for weldability with
82
Expansion Bi = BuoH
l
Bj = 1.7 T
Deflection of magnetostriction
pertinent to peak Bj of the magnetic
polarization, l as a function of
mechanical stress o, both valus in
rolling direction
Transformers, 2/e
–10 –8 –6 –4 –2 2 4 6 8 10
Compression Contracting Tension
MPa
Contraction
Figure 3.17 Alternating field magnetostriction CRGO as a function of external mechanical stress.
standard quality magnetic steels. Depending upon the requirements

of application of the component, particular quality of steel is used,
details of which are given in Chapter 10.
Mechanical properties of a few qualities of structural steel are
Table 3.12 Properties of Structural Steel
Sl. Material Tensile Yield Elongation

No. strength strength %, (Min.)
MPa MPa (Min.)
1. Cold rolled carbon steel sheet 275 — —
2. Hot rolled carbon steel sheet 330–410 205 18 (on 80 mm
gauge length)
25 for thick-
ness above
3 mm.
3. Structural steel Std. quality
Nominal thickness/diameter, mm
(i) 6 to 20 410–530 250 23
(ii) 20 to 40 410–530 240 23
(iii) Over 40 410–530 230 23
4. Bright steel bar and section-cold 430 (min.) 335 14
drawn
5. High tensile strength structural
steel plate
Nominal thickness/diameter (mm)
(i) 6 to 16 540 (min.) 350 20
(ii) 16 to 32 540 (min.) 340 20
(iii) 32 to 63 510 (min.) 330 20
(iv) Above 63 490 (min.) 280 20
6. HTS Steel Bar (1.5% nickel- 900–1050 700 15
chromium-molybdenum)-
hardened and tempered
7. Austenitic chromium-nickel-steel- 520 (min.) — 40
titanium stabilized plate
(stainless steel)
3.10 Future Trends
In the recent years, intensive efforts have been undertaken to

develop new transformer insulation systems which permit a more
compact and thus more economical design. However, these have
been, more or less, limited to small transformers up to a few MVA.
Dry type transformer, silicone oil filled transformer, the vapour
cooled and SF6-insulated transformers are already available today.
However these special transformers are used for a specific purpose,
e.g. for weight reasons in locomotives and are normally costlier
than conventional oil-cellulose transformer. The current level of
knowledge and higher cost put limitations on these new insulation
systems. Studies have shown that conventional oil barrier insulation
system will cause no problems up to operating voltages of 2000 kV.
Hence, it is very unlikely that oil impregnated cellulose insulation
system will be replaced by a new insulation system for large size
power transformer, at least in the near future.
To increase the short-circuit strength of windings, annealed
copper conductor is already being replaced by controlled proof stress
copper. Copper-silver alloy will replace high conductivity copper,
specially in higher rating transformers. Under short-circuit
conditions in the windings, copper-silver alloy conductor is less
susceptible to annealing and is thus easily able to retain its strength.
Mechanical strength of transposed conductor is further increased by
bonding together all the conductors in a stack, by use of an epoxy
bonding enamel over PVA enamelled strips. These changes are likely
to take place faster, to cover almost all transformers. However,
copper is expected to be used unless it is replaced by a new generation
development of a superconducting material, since anodized alu-
minium strip conductors have their limitations for HV windings.
Another possible substitute for grain-oriented electrical steel for
magnetic core of transformer is amorphous steel, also referred as
metallic glass, which is a non-crystalline solid created by rapid
quenching of metal-metalloid alloys. Amorphous steel (typical com-
position Fe8t B13.5 Si3.5 C2) of thickness 25 to 50 mm and widths up to
100 mm are now commercially available. It is expected that 175 mm
wide strips would be commercially available shortly. However,
there is apprehension about the technological capability of produc-
ing metallic glass in widths of 800–1000 mm to match with CRGO
width. There are numerous practical problems to be sorted out, e.g.

complete transformer will have to be annealed for stress relieving,
therefore pre-wound transformer coils are to be made suitable to
withstand annealing temperature. Further, saturation induction of
metallic glass is lower than that of CRGO and hence transformer
has to be designed at lower induction. Experimental transformers
of maximum rating of 100 KVA have been made so far, using
amorphous steel. Nevertheless, the material holds some promise
for the future, as it is expected that the problem due to its present
prohibitive cost and other limitations can be overcome.
REFERENCES
1. Viswanathan, P.N. et al.,
“Paper for Cable Insulation from Indigenous Wood Pulps”, National
Seminar on Electrical Insulation held in Nov. 1974 sponsored by the
Institution of Engineers (India), Electrical Engg. Division, IEMA
and ISI.
2. Philip, P.K., “Insulation Pressboard for Transformers”, Electrical
India, Vol. XIX, No. 24, pp. 5–13, Jan. 1980.
3. BICC Connolly’s U.K. Publication No. 798
‘Continuously Transposed Conductor’.
4. Thyssen Grillo Funke, W.G., Brochure, “Grain Oriented Electrical
Sheet”, 1980 edition.
5. “Technical Data on Orient-Core, HI-B”
Nippon Steel Corporation, Japan, Cat. No. EXE 367, Dec. 1976.
6. “Technical Data on Domain Refined Orient Core HI–B. Nippon Steel
Corporation, Japan. Cat. No. EXE. 706, Nov. 1987.
C HAPTER 4
Magnetic Circuit
K. N. Labh, R.C. Agarwal
In a transformer, energy is transferred from one electrical circuit

to another through the magnetic field. Transformer core made of
laminated sheets provides the magnetic circuit for the flow of mag-
netic flux mutually linking the electrical circuits. As against the air
core, iron core provides a comparatively low reluctance path to the
magnetic flux with consequent benefit of (a) smaller magnetizing
current, (b) increase in the total flux linkage and (c) a high ratio of
mutual to leakage flux resulting in reduction of stray losses. Its
design, type and manufacturing methods have significant bearing
on quality, transportability, operational limitations and guaranteed
technical performance of the transformer.
4.1 Material
Some of the very early transformer cores were made of inferior

grades of laminated steels which had inherently higher core losses
and showed pronounced ageing effects, further aggravating the
hysteresis component of iron-losses in the equipment. It was subse-
quently found that very small quantities of silicon alloyed with low
carbon content steel produced a material with low hysteresis losses
and high permeability. These steel sheets alloyed with silicon
mitigated the problem of ageing and improved the permeability and
consequently reducing the magnetizing current and core losses. In
the ever increasing pursuit of increasing the power ratings and
reduction of core-losses, another innovative technique from steel
manufacturers came in the form of cold rolling with orientation of
the grain in the direction of rolling. This core steel known as CRGOS
Magnetic Circuit 87
(cold rolled grain oriented silicon steel) has the minimum epstein
losses to the flow of magnetic flux along the directions of grain ori-
entation and this material is universally used for the manufacture
of transformer cores.
The adoption of CRGOS has brought about considerable reduc-
tion in the specific iron-losses (W/kg) over the earlier grades of core
steel. However, CRGOS is susceptible to increased losses due to flux
flow in directions other than that of grain orientation, effect of
mechanical strain due to clamping pressure, bolt holes, jointing of
limb with yokes,1 etc. Apart from this sensitivity to the direction
of rolling, CRGOS sheets are also very susceptible to impaired
performance due to impact of bending, blanking the cutting. Both
surfaces of the core steel sheets are provided with an insulating
of oxide coating (commercially known as Carlite). The stacking
factor of lamination improves by using thicker laminations, but
eddy current loss goes up in proportion to square of the thickness
of the lamination. For reducing the eddy current losses, thinner
laminations are preferable even though the stacking factor goes
down. Deburring of the laminations improves the stacking factor
and minimizes the eddy losses. After machining, the material has
to be annealed at 800–900°C in a neutral gas environment. The
material properties of core steel are further discussed in Chapter 3.
4.2 Design of Magnetic Circuit
For a transformer design, the basic governing factors are:

(a) Rating of transformer and its performance,
(b) Operational conditions,
(c) Transport limitations (i.e. height, length, width and weight,
etc.)
The design of the magnetic circuit, i.e. transformer core is also
based on the above considerations and it has significant bearing on
the overall economy of the transformer. For CRGOS, saturation
may occur at the magnetic flux densities exceeding 1.9 tesla. Based
on the input voltage and frequency variations, a suitable value of
flux density can be adopted to avoid any chance of core saturation
under operating conditions. By increasing the operating value of
magnetic flux density (magnetic loading) the net weight of core can
be reduced, but this leads to corresponding increase in the core
losses. One has to compute an optimum value of the magnetic flux
density, keeping in view all the above considerations.
4.2.1 Constructional Features
The type of transformer core construction depends on the technical

particulars of the transformer and transport considerations. In
general it is preferable to accommodate the windings of all the
three phases in a single core frame. Three-phase transformers are
economical over a bank of three single-phase transformers. Another
important advantage of three-phase transformer cores is that
component of the third and its multiple harmonics of mmf cancel
each other, consequently the secondary voltage wave shape are free
from distortions due to the third harmonics in mmf. However, if the
three-phase ratings are large enough and difficult to transport, one
has no choice but to go for single-phase transformer units.
For single-phase and three-phase transformers, the cores can be
broadly classed as:
(a) Single-phase three-limbed core
(b) Single-phase two-limbed core
(c) Three-phase three-limbed core
(d) Three-phase five-limbed core.
(a) Single-phase Three-limbed Core Fig. 4.1(a)

The windings are placed around the central limb, also known as
main limb. The main magnetic flux generated in the central limb
gets divided into two parallel return paths provided by the yokes
and auxiliary limbs. For the same magnetic flux density as that in
the main limb, the auxiliary limbs and the yokes need to have the
cross section only half of the main limb. This type of transformer
core is generally preferred for single-phase transformer, as this is
more economical than two limbed construction discussed below.
(b) Single-phase Two-limbed Core [Fig. 4.1(b)]

Sometimes the single-phase power ratings of transformers are so
large that if the windings of full power ratings were to be placed on
the central limb, its width would become too large to be transported.
Magnetic Circuit 89
1
2
Figure 4.1 (a) Single-phase three-limbed core. (1) Main limb

(2) Top yoke (3) Bottom yoke (4) Aux. limb
Figure 4.1 (b) Single-phase two-limbed core. (1) Main limb

(2) Top yoke (3) Bottom yoke
To mitigate such difficulties the windings are split into two parts
and placed around two separate limbs. Here the cross-sectional area
of the legs (limbs) and the yokes are identical. Consequently these
cores are bulkier than the single-phase three-limbed arrangements.

Also the percentage leakage reactance for this type of core construc-
tion is comparatively higher due to distributed nature of the
windings in the two limbs separately.
(c) Three-phase Three-limbed Cores (Fig. 4.2)

This type of core is generally used for three-phase power transformer
of small and medium power ratings. Each phase of the winding is
placed around one leg. For each phase of magnetic flux appearing in
a limb, the yokes and the remaining two limbs provide the return
path. If the phase fluxes are denoted as fA, fB, fC, their summation at
any instant of time is identically zero, which can be mathematically
stated as fA + fB + fC = 0. In this type of construction, all the legs
and the yokes have identical cross section.
Figure 4.2 Three-phase three-limbed core: (1) main limbs,

(2) top yokes, (3) bottom yoke.
(d) Three-phase Five-limbed Cores (Fig. 4.3)

For large rating power transformers, cores have to be built in large
diameters. In case of three-phase three-limbed cores, the yokes
Magnetic Circuit 91
1 2
4 4
6 6
5
3
Figure 4.3 Three-phase five-limbed core: (1) main limbs, (2) top
main yokes, (3) bottom main yokes, (4) auxiliary top
yokes, (5) auxiliary bottom yokes, (6) auxiliary
limbs.
have the same diameter as the limbs. In case of large diameter

cores, the overall core height will go up leading to transport problem.
For such cases the yoke cross-sections (and consequently yoke
heights) are reduced by approximately 40% or more and auxiliary
paths for the magnetic flux are provided through auxiliary yokes
and limbs. The cross-section and the height of the auxiliary yokes
and limbs are lower than that of the main yokes.
4.3 Optimum Design of Core
For the optimum design of magnetic core, the following aspects have
to be decided.
(a) Constructional features
(b) Core cross-sectional area
(c) Number of oil ducts and location
4.3.1 Core Cross-Section
The ideal shape for the section of the cores is a circle, since this
would waste no space beyond that taken up by the insulation
between laminations. A perfectly circular core section, however,

involves making a variation in dimensions for each successive
lamination, which is possible but uneconomical. As a compromise
solution, the core cross-section is made by laminations of varying
widths and packet heights in such a way that the overall section
approximates a circle. Such a typical core cross-section is shown in
Fig. 4.4. Oil ducts are needed for cooling the core, lest the hot-spot
temperature rises dangerously and their number depends on the
core diameter and the specific core-losses, which is a function of
operating flux density. Additionally, clamp plates made of steel are
needed on either side of laminations for effectively clamping the
laminations. These clamp plates should be mechanically strong
enough to prevent buckling/bending of laminations and be able
to withstand the lifting load of core and windings and axial short-
circuit forces. The steel sheet laminations, oil ducts and the clamp
plates should all lie within the core circle.
2 H
Figure 4.4 Circular core cross-section: D—diameter of core,

H—total lamination stack height (1) laminations,
(2) oil duct, (3) steel clamp plates.
Magnetic Circuit 93
The net sectional area is calculated from the dimensions of

various packets and an allowance is made for the space lost between
laminations (known as stacking factor) which for sheet steel of 0.28
mm thickness with carlite insulation coatings is approximately
0.96. Area is also deducted for the oil ducts. The ratio of the net
cross-sectional area and the gross area of the core circle known as
utilization factor (UF). By increasing the number of core steps UF
improves. This, however, also increases the manufacturing cost.
Typical cost effective values for the number steps (i.e. one-half
of cross-section from centre line) lie in the range of 6 (for smaller
diameters) to 15 (for large diameters). For any particular core
diameter based on other design considerations, this gives out not
only the optimum area and thereby reduction in the flux density
and consequently iron-loss, but also helps the designer to revert to
lesser value of core diameter, wherever the computational margins
allow this latitude. Seen from another angle, improvement in the
core utilization factor increases the core area and hence the value of
volts/turn for any particular core diameter and specified flux
density. This, in turn, results in the reduction in winding turns and
thus reduction in copper. Therefore, core area optimization results
in better economy of transformer designs.
In the following, the core optimization is discussed under (a) opti-
mum selection of laminations (b) optimum oil ducts.
4.3.2 Optimum Selection of Laminations
For any particular core diameter, the first and foremost point to be
decided is the maximum allowable height of the lamination packets.
This, in turn, is determined by the design consideration of clamp
plate (see Fig. 4.5) and the pertinent constraint in accommodating
them inside the core circle.
For a circle of diameter D, the length of the cord at distance Y
from the centre is given by
F F DI 2 I
F(Y) = 2 GG GH 2 JK - Y2 JJ (4.1)
H K
If H is the maximum allowable packet height, the minimum
allowable lamination width is given by
F (Y )
Figure 4.5 Circular cross-section of core-dimensional relationship.
F F DI F H I
2 2 I
Lmin. = 2 GG GH 2 JK - GH 2 JK JJ (4.2)
H K
The maximum allowable width of the central packet is usually core
diameter less G mm (to accommodate wooden packing)
i.e. Lmax. = D – G (4.3)
If the core is to be built up in Ns number of steps, where widths of
laminations of individual steps are Li, i = 1, 2,... Ns, these values
must satisfy the following relations
Lmin. £ Li £ Lmax. (4.4)
i = 1, 2,...., Ns
For building the circular cross-section of the core in say Ns steps, we
have to decide Ns different widths of laminations to be stacked one
over the other as shown in Fig. 4.4. The packet height of each
lamination is computed by the difference of heights of cords equal
to the width of laminations and the adjacent next width below it.
The gross cross-sectional area contributed by individual lamination
Magnetic Circuit 95
packets is obtained by lamination widths times packet height. This

is taken as the return function for selection of a particular width.
The optimum selection of laminations is formulated on the method
of dynamic programming. Recursive relationships are formulated
for return function and state transition function for each step of
decision taking. This method gives optimum selection of laminations
for building in Ns number of steps.
Out of the optimum area available by the above method we must
deduct areas for the oil ducts. The individual packet heights should
be reduced slightly to allow for the overall manufacturing tolerance
in the core built up. A computer program has been developed for
the optimum selection of lamination packets based on the above
algorithm. The computer program also computes (a) optimum
number of oil ducts and their locations and (b) core lamination
dimensions and other details required for manufacturing activities.
This has automated the complete transformer core design and
manufacturing informations.
4.3.3 Computation of Optimum Number of Oil Ducts

Core-losses (iron-losses) take place due to magnetic flux flow in the
laminations. For CRGOS these losses are minimum for flux along
the direction of grains orientation and maximum for flux flow along
cross grain directions. Hot temperature spots are developed inside
the core as a result of core-losses and designers must ensure that
the hot-spot temperatures are well below the permissible values.
Segments of the core cross-section are rectangular as shown
in Fig. 4.6. Heat dissipation takes place along the laminations
(x-direction) and across the laminations (y-direction). The tempera-
ture gradients along the two directions are given by
Tx = W◊x◊
FG x +h
IJ (4.5)
H 2K 1 K
Ty = W◊y◊ G
F y + hJ
I (4.6)
H 2K 2 K
where
W = specific cross gain loss, W/kg,
K1 = thermal conductivity along the laminations, W/°C/
mm,
K2 = thermal conductivity across the laminations W/°C/
mm,
Heat flow across laminations

Hot spot
Y-DIR
Heat flow along laminations
2Y
X-DIR
2X
Figure 4.6 Rectangular stack of core laminations heat flow conditions.
h = surface heat transfer coefficient, °C/mm2/W,

Tx = temperature differential from hot spot to oil, assuming
that the all heat flows along the laminations.
Ty = temperature differential from hot spot to oil across
lamination, assuming all heat flows across lamination.
The hot-spot temperature gradient
Tx ◊ Ty
Th = (4.7)
Tx + Ty
Th (max. permissible) = maximum permissible hot-spot tempera-
ture–(oil rise + ambient temperature)
Equation (4.7) can be rewritten as
Tx ◊ Th
Ty = (4.8.1)
Tx - Th
i.e. W◊y
FG y IJ
+h =
Tx ◊ Th
(4.8.2)
H 2K 2 K Tx - Th
The value of W depends on the type of core construction and
remains constant for a particular type of assembly. The value of Tx
increases parabolically with the value of x. An interesting condition
arises when x = x* such that
Tx £ Th (4.9)
The equality condition of (4.9) implies that y = •, i.e. for x = x*, Th
will never achieve its specified maximum value for value of y, and,
Magnetic Circuit 97
along the lamination, dissipation of heat will be sufficient to

maintain the value of Th below the specified maximum permissible
value. The second condition that Tx < Th leads to negative value of
Ty, which is not physically possible. This also implies that along the
lamination, dissipation is sufficient to contain the value of Th below
the specified limit. For such conditions as obtainable in relation
(4.9), no oil duct is required for cooling.
We now examine the case when Tx > Th. Equation (4.8.2) can be
solved for y as
1
y* = ( - h + h 2 + 4C1 ◊ C2 ) (4.10)
2C1
1
where C1 =
2K2
U|
|V
and C2
F T ◊ T IJ W |
=G x h
(4.11)
H T - T K |W
x h
Here, we introduce the concept of critical height y £ y* obtainable

from Eq. (4.10). For any stack height 2y such that y £ y* for a specified
value of x, the hot-spot temperature will never exceed the specified
value of Th. For actual stack height H > 2y*, cooling surface (in the
form of oil duct) is required if the hot-spot temperature is not to
exceed Th.
For any core diameter D, the minimum number of oil ducts
is calculated by iteratively using the Eqs. (4.10) and (4.11) for
computation of critical heights y* starting from the central packet.
Since above formulation stipulates that the maximum hot-spot
temperature in the packet will be less than or equal to the specified
permissible value, the above calculation gives the optimum number
of oil ducts.
4.3.4 Location of Oil Ducts

For more than one oil duct, a suitable criterion is required regarding
their location in the core section. For single duct, the obvious
choice is its location at the centre. Minimum difference of hot-spot
temperatures (ideally zero) obtaining in different regions divided
by the oil ducts is adopted here as the criterion of location of oil
ducts. Figure 4.7 represents the case of two oil ducts in the core
cross-section.
Ducts
Zone 2
Zone 1
2x 1 2y 1
Figure 4.7 Two oil ducts in core.
For Zone 1
1 1 1
= + (4.12)
Th1 Tx 1 Ty1
Where Tx1 = W◊x1

FG x 1
+h
IJ U|
H 2K 1 K |V
Ty1 = W◊y1
FG y 1 I|
+ hJ | (4.13)
H 2K 2 KW
and Th1 = Hot-spot temperature gradient of zone 1.
For Zone 2
1 1 1
= + (4.14)
Th2 Tx 2 Ty2
where Tx2 = W◊x2

FG x 2 IJ U|
+h (4.15)
H 2K 1K| (4.15)
Fy I V|
2
Ty2 = W◊y G + hJ
2
H 2K 2K |W
and Th2 = Hot-spot temperature gradient of Zone 2.
Magnetic Circuit 99
By equating the hot-spot temperature gradients of zones 1 and 2, we

obtain the following:
2 K1 2K 2
+
W ◊ x1 ( x1 + 2 K 1h) W ◊ y1 ( y2 + 2 K 2 h)
2K2 2 K2
= + (4.16)
W ◊ x2 ( x2 + 2 K 1h) W ◊ y2 ( y2 + 2 K 2 h)
By cancelling the common terms and expressing the difference of
expressions of l.h.s. and r.h.s. alongwith a normalizing constant we
compute a difference term as
C0 = p
FG 1
-
1
Hx 1 ( x1 + 2 K 1 h) x2 ( x2 + 2 K 2 h)
+
K2
-
K2 IJ
y1 ( y1 + 2 K 1 h) y2 ( y2 + 2 K 2 h) K
(4.17)
where K2 = K2/K1 UV
and p = normalizing constant (4.18)
y1 + 2y2 = H/2
W
2
U|
F DI
x = G J -y
2
2 V| (4.19)
H 2K 1
W
Since x1 is a constant and y and x are related to x and y Eq. (4.19),
2 2
we have
C0 = f(y1) (4.20)
The desired value of y is the one which satisfies C0 ~
- 0. This is
obtained by an iterative procedure on computer. The distance of oil
ducts from centre of core circle are similarly computed for number of
oil ducts more than two no similar principle.
In the foregoing, the broad concepts of optimum design of trans-
former cores of circular cross-section has been discussed. Based on
the optimum core cross-sectional area and operating flux density,
the windings for individual phases can be designed and from this
the core window size can be worked out. This decides the overall
transformer core frame dimensions for a particular type of core
construction as illustrated in Figs. 4.1, 4.2 and 4.3.
4.4 Manufacturing
By adopting suitable technology for transformer cores and exercising

care in handling of laminations and core building, it is possible to
achieve
(a) higher reliability
(b) reduction in iron-losses and megnetizing current
(c) lowering material and labour cost
(d) abatement of noise levels
It is necessary to apply quality checks at different stages of manu-
facturing to ensure quality and reliability. The core steel samples
should be tested regularly for guaranteed epstein iron-loss values.
The laminations should also be visually inspected and the rusted lot
should be rejected. CRGOS sheet steels are susceptible to impaired
losses due to cutting, punching, piercing, bending, etc. During cutting
and piercing the edges develop burrs which may cut, as knife-edge,
the insulation coatings on the adjacent laminations, in addition to
lowering the stacking factor. It is imperative that laminations are
deburred and annealed (for stress relieving), so that iron losses do
not increase. For reducing the transformer noises, the laminations
should be tightly clamped together and punch holes should be
avoided as far as possible. The air gap at the joints can be controlled
by working on tight tolerances so that value of magnetizing
currents are kept to a minimum. These manufacturing aspects are
discussed in more detail in the following:
4.4.1 Corner Jointing of Limbs with Yokes

Broadly speaking, the core losses can be split into (a) loss due to
magnetic flux flow along the direction of grain orientation (with grain
iron-losses) and (b) flux flow in cross-grain direction (cross-grain
losses) occurring in the zones of jointing of limbs with yokes. The
cross-grain losses depend to a large extent on the type of joints. The
two must commonly used types of corner joints are (a) interleaved,
(b) mitred.
(a) Interleaved Joints (Fig. 4.8)
Interleaved joints are the simplest, from the point of view of manu-
facturing. However, in the cross-grain zones the magnetic fluxes
leave/enter the laminations in perpendicular direction to the grains
and these losses are comparatively higher. Such type of jointings
are usually preferred only for small rating transformers, where the
total core-loss itself is very small.
(b) Mitred Joints (Figs 4.9, 4.10 and 4.11)
When the corners of the laminations are cut at 45°, the jointing
is known as mitred joint. The cross-grain losses for this type of
Magnetic Circuit 101
Figure 4.8 Three-phase three-limbed core-interleaved joints
Figure 4.9 Three-phase three-limbed core mitred joints.
jointing is minimum, as the magnetic flux leaving/entering at the

joint finds a smooth path of its flow. This, however, entails extra
manufacturing cost for preparing the corner edges of the individual
laminations.
Figure 4.9 shows one arrangement of laying the laminations for
three-phase three-limbed cores with mitred joints. Figure 4.10 shows
the laminations laying arrangement for the three-phase five-limbed
cores with mitred joints.
Sometimes there may be constraints in preparation of laminations
beyond certain widths. If lamination widths larger than these are
required, as in case of large diameter cores, these are split into two
halves, so that these can be handled easily in manufacturing. A
typical core built with such split laminations is shown in Fig. 4.11.
Figure 4.10 Three-phase five-limbed core mitred joints.
Figure 4.11 Three-phase five-limbed core with mitred joints with

split laminations for large core diameters.
4.4.2 Preparation of Lamination Sheets

(a) Slitting
For building the transformer cores, lamination sheets of different
widths and packet heights are needed. The manufacturing schedule
may include cores of different diameters and different types of
constructions necessitating slitting laminations in many widths
and lengths. CRGOS rolls cannot be ordered in so many different
widths and quantities. These rolls are available in standard widths
of say 760, 790, 840, 915, 1000 mm, etc. For slitting operation, some
widths can be combined together by suitably adjusting the cutter

distances in the slitting machine. Figure 4.12 (Plate 2) shows slitting
operation of laminations.
It is evident that full width of roll cannot be utilized at any time
of slitting operation and the leftover material will vary from stage
to stage and depending on the widths selected in combination
during the process of slitting. The meticulous care in planning is
imperative to minimize wastage of core steel.
The slitting operation has been formulated based on the principle
of dynamic programming to compute the optimum schedule. This
provides the optimum combination of different widths to be taken
together and the length for slitting, such that the scrap during
slitting operation is minimum.
Figure 4.13 (Plate 2) shows the slitted rolls of different widths,
which are collected at the other end of the slitting operation over a
collapsible mandrel.
(b) Cropping of Laminations

The different shapes and sizes of laminations needed for core building
are illustrated in Figs. 4.8–4.11 for different types of transformer
cores. Figure 4.14 (Plate 2) shows a cropping machine for cutting
laminations from slitted rolls.
In case of mitred core laminations, these are first cut in a
trapezoidal shape. Finished shape is given by cutting off the corners
(wherever needed) at the second stage of cutting by employing
simple hydraulically operated guillotine machines.
(c) Piercing Operation

The yoke punchings (laminations) usually need holes for bolting the
yoke laminations. These holes are punched after the cropping/guil-
lotine operation by suitably adjusting the hole piercing positions in
the piercing machine and selecting the right piercing tool for speci-
fied hole punch size. Figure 4.15 (Plate 2) shows a piercing machine.
However, some cores are also constructed without yoke bolts.
(d) Deburring
During the process of slitting, cutting and piercing of laminations,
the cut edges get some burrs. These burrs are removed by passing
the laminations through deburring operations. Presence of burrs
impairs the stacking factor. Also burrs cut into the insulation
coatings and bridge adjacent laminations and thereby increase the
eddy losses.
(e) Annealing/Varnishing
If the insulation coatings at the edges are scratched during deburring
process or extra varnish coating is desired, the laminations are
processed in a varnishing plant which provides a thin coating of
varnish and quickly dried up at elevated temperature. However,
varnish coating is not considered as necessary if the carlite insula-
tion coating on the lamination is consistent. During the process of
slitting, cropping, piercing, deburring, etc., mechanical strains are
developed inside the laminations, which disturb the original grain
orientation and thereby increase the iron-losses. This problem
is mitigated by annealing the laminations in an annealing plant
[Fig. 4.16 (Plate 2)].
Annealing is done at an elevated temperature of 800–900°C,
preferably in a neutral atmospheric zone and subsequently cooled
by a blast of air.
The finished laminations are then taken to the core assembly
area.
4.4.3 Core Assembly

Core building from the finished lamination sheets is done in
horizontal position on specially raised platforms. The lamination
sheets are susceptible to mechanical stresses of bending, twisting,
impact, etc. A lot of care is exercised while handling and normally
two persons are needed to hold the two ends of the laminations
at the time of laying. Figure 4.17 (Plate 2) shows the core building
operations typically for a three-phase five-limbed core.
At first the clamp plates and end frame structure of one side of the
core assembly are laid out. Guide pins are used at suitable positions
for maintaining the proper alignments during core building process.
Oil ducts are formed by sticking strips on lamination and put in
position as required.
For each packet, the laminations are manufactured in two differ-
ent lengths and these sets are laid out alternately, keeping at a time
two to four laminations together. The two alternate arrangements
provide overlapping at the corner joints and when the lamination
packets are clamped together, these overlapping edges provide
sufficient mechanical strength in holding the edges in tight grip.
After laying out the complete laminations, the clamp plates, and
end frame structure of the other side are laid out and the entire
core-end frame structure is properly secured through bolts and steel
bands at a number of positions.
The platform on which the core building takes place is of special

design and the core-end frame assembly can be raised to the vertical
position alongwith the platform which serves as a cradle. Subse-
quently the platform is disengaged. In this process, the core assembly
is spared from the mechanical strain of lifting and raising in the
vertical position. Small-size cores can however be built up without
these special platforms.
Steel bands used for tightening the laminations is only a tempo-
rary arrangement and are later removed, otherwise these will form
short circuited turns. Two commonly used methods of holding the
leg laminations together is their clamping by either (a) resiglass
tape or (b) using skin stressed bakelite cylinders. In case of resiglass
tapes, these are tightly wound around the legs at specified pitch and
cured by heating. The tape shrinks after heating and provides a
firm grip. The tensile strength of resiglass tapes is even higher than
that of steel tapes. In the case of core legs tightened by skin stressed
cylinders (base cylinder of innermost coil), these are lowered from
the top and the steel bands and cut progressively. Wooden wedges
are inserted along the packet corners and hammered down, so that
the enveloping bakelite cylinder and the leg laminations are fitting
tightly against each other. Figures 4.18 (Plate 3) and 4.19 show the
complete core-end frame assemblies for single-phase three-limbed
and three-phase five-limbed cores.
Conventionally, the core is assembled along with all the yokes,
and after assembly the top yokes are unlaced after removing the
top-end frames for the purpose of lowering the windings. This takes
a lot of labour and manufacturing time. The latest development is to
assemble the core without top yokes and insert the top yokes after
lowering all the windings in the core leg.
4.4.4 Fitting of Core in the Tank

The most commonly used method of putting the core assembly is to
rest the core frame on its feet, which in turn is firmly fixed on the
tank base. The extreme end feet are enclosed in a steel bracket
welded on the tank base. The top portion of core assembly is also
suitably locked with the tank cover, so that any possible magnifica-
tion of the vibration during transit is fully arrested. An alternative
to this is construction of the tank base in the shape of a channel. The
laminations along with the bottom-end frames are jacket from
either side against the channel walls of the tank. As such, no bolts are
needed in the bottom yokes and iron-losses are comparatively lower
because of the absence of these punch holes. Figure 4.20 illustrates

this channel type of tank and core assembly fixing arrangement.
This also has the advantage that the core is rigidly fixed to the tank
bottom and no possibility of core assembly shifting exists. As a re-
sult of raising the tank base (Item 1, Fig. 4.20) the inner oil volume
inside the tank is reduced. Also, the windings and the insulation
rings and blocks are directly resting on the flat tank base.
6 6
7 7
5
4
3
Figure 4.20 Transformer core fitting in channel-shaped tank base

1. Channel-shaped tank base
2. End view of transformer core
3. Bottom end frame
4. Jacking of core end frame against channel walls
5. Tank base stiffener
6. Windings
7. Insulation blocks and rims
REFERENCES
1. Brechha, H., “Some Aspects of Modern Transformer Core Design”,
Bulletin Oerlikon, No. 324, pp. 70–80.
2. Brechna, H. “New Design Trends in Construction of Transformer
Cores”, Bulletin Oerlikon, No. 326, pp. 6–14.
3. Austen Stignant and Franklin, J and P Transformer Book,
Newness–Butterworths, 10th Edition, 1973.
4. Bhusan Prabhakar, “Optimum Design of Distribution Trans-
formers”, M. Tech. Thesis, 1971, Deptt. of Electrical Engg., IIT
Kanpur.
C HAPTER 5
Windings
and Insulation
M. V. Prabhakar
S.K. Gupta
Windings form the electrical circuit of a transformer. Their construc-

tion should ensure safety under normal and faulty conditions. The
windings must be electrically and mechanically strong to withstand
both over-voltages under transient surges, and mechanical stress
during short circuit, and should not attain temperatures beyond the
limit under rated and over-load conditions. For core-type transform-
ers, the windings are cylindrical, and are arranged concentrically.
Circular coils offer the greatest resistance to the radial component
of electromagnetic forces, since this is the shape which any coil will
tend to assume under short circuit stresses.
5.1 Types of Windings
The choice of the type of winding is largely determined by the rating

of the winding. Some of the common types of windings are described
below.
5.1.1 Distributed Cross-over Windings
These windings are suitable for currents not exceeding about 20 A.

They comprise wires of circular cross-section (Fig. 5.1) and are used
for HV windings in small transformers in the distribution range. A
number of such coils are joined in series, spaced with blocks which
provide insulation as well as duct for cooling.
Windings and Insulation 109
Figure 5.1 Cross-over coils.
5.1.2 Spiral Winding
This type of winding is normally used up to 33 kV and low current

ratings. Strip conductors are wound closely in the axial direction
without any radial ducts between turns. Spiral coils are normally
wound on a bakelite or pressboard cylinder (Fig. 5.2).
Bakelite cylinder
Figure 5.2 Spiral coil (single layer, wound on flat side).
Though normally the conductors are wound on the flat side,

sometimes they are wound on the edge. However, the thickness of
the conductor should be sufficient compared to its width, so that the
winding remains twist-free (Fig. 5.3).
Figure 5.3 Spiral coil (edge wound).

Spiral windings may be made as single layer or multilayer type.

Figure 5.4 shows a double-layer spiral coil where an oil duct sepa-
rates the two layers. For such a coil, both the start and the finish
leads lie at one end of the coil and may at times prove to be advanta-
geous for making the terminal gear.
Figure 5.4 Spiral coil (double layer).
Normally it is not necessary to provide any transposition between

the parallel conductors of a spiral winding as the lengths and the
embracing of leakage flux are almost identical.
5.1.3 Helical Winding
This type of winding is used in low-voltage and high-current ratings.

A number of conductors are used in parallel to form one turn. The
turns are wound in a helix along the axial direction and each turn is
separated from the next by a duct. Helical coils may be single-layer
(Fig. 5.5) or double layer (Fig. 5.6) or multi-layer, if the number of
turns are more.
Unless transposed, the conductors within a coil do not have the
same length and same flux embracing and therefore have unequal
impedance, resulting in eddy losses due to circulating current
between the conductors in parallel. To reduce these eddy losses, the
helical windings are provided with transposition of the conductors
which equalise the impedances of the parallel conductors.
5.1.4 Continuous Disc Winding
This type of winding is used for voltage between 33 and 132 kV and
medium current ratings. These coils consist of a number of sections
Start Finish
Figure 5.5 Helical coil Figure 5.6 Helical coil

(single layer). (double layer).
placed in the axial direction (Fig. 5.7), with ducts between them.
Each section is a flat coil, having more than one turn, while each
turn itself may comprise one or more conductors (usually not more
than four or five), in parallel. The sections are connected in series,
but without any joints between them. This is achieved by a special
method of winding. It is not necessary to provide a cylindrical
former for these coils, as these are self-supporting. Each disc is
mechanically strong and exhibits good withstand of axial forces.
Another particular advantage of these coils is that, each section can
have either integral or fractional number of turns (for example
15
4 turns per section).
18
Figure 5.7 Continuous disc winding.

5.1.5 Transposition
(a) For helical windings, usually three transpositions are

provided. The complete transposition [Fig. 5.8(a)] is provided
in the middle of the windings. Two partial transpositions are
provided, one at 25% of turns [Fig. 5.8(b)] and the other at
75% of turns [Fig. 5.8(c)]. In complete transposition, each
conductor position is varied symmetrically, relative to the
middle point, whereas in partial transpositions, two halves
of parallel conductors are interchanged in the positions: the
upper half becomes the lower, and vice versa. Such a
transposition needs extra space in the height of the coil.
6 3 4 1
5 2 5 2
4 1 6 3
3 6 1 4
2 5 2 5
1 4 3 6
(b) (a) (c)
Figure 5.8 Transpositions in helical winding (a) Complete

transposition (b) and (c) Partial transposition.
(b) With a multi-start helical winding, the transposition can be

achieved by using rotary transposition. Figure 5.9 shows
transposition in a double-start helical winding. By this
arrangement, every conductor occupies every position by turn
and thereby complete equalization of impedance is possible.
Also, there is no need for extra space in the coil height.
1 12 2 1 3 2 4 3 12 11
2 11 3 12 4 1 5 2 1 10
3 10 4 11 5 12 6 1 2 9
4 9 5 10 6 11 7 12 3 8
5 8 6 9 7 10 8 11 4 7
6 7 7 8 8 9 9 10 5 6
Figure 5.9 Rotary transposition for double-start helical winding.

(c) For disc windings having more than one conductor in parallel,
transposition is made between the conductors by changing
their mutual position at each cross-over from one section to
another (Fig. 5.10).
1 2 3
1 2 3
Figure 5.10 Transposition at each cross-over in continuous disc winding.
5.1.6 Interleaved Disc Winding

A disadvantage with the continuous disc winding is that their
strength against impulse voltages is not adequate for voltages
above, say, 145 kV class. The impulse voltage withstand behaviour
of disc coils can be increased if the turns are interleaved in such a
fashion that two adjacent conductors belong to two different turns.
Figure 5.11 shows such a winding in which interleaving has been
done in each pair of discs. It will be noticed that it is necessary to
have 2n conductors in hand for winding when n is the number of
12 4 11 3 10 2 9 1
Interleaved
disc pair
5 13 6 14 7 15 8 16
18 17
Figure 5.11 Interleaved disc winding (2 discs per group).

conductors in parallel. Conductors of turns 8 and 9 are joined by

brazing. A cross-over is given at the bottom of the disc.
Apart from interleaving between every double-disc, it is also pos-
sible to have more number of discs (say four) in each interleaved
group (Fig. 5.12).
20 4 19 3 18 2 17 1
5 21 6 22 7 23 8 24
Interleaved
disc-quad.
28 12 27 11 26 10 25 9
13 29 14 30 15 31 16 32
34 33
Figure 5.12 Interleaved disc winding (4 discs per group).
This gives further improved behaviour against impulse voltage,

though there are concomitant increased complexities.
Interleaved windings require more skill and labour than plain
continuous disc windings. Sometimes a part of the winding is inter-
leaved while the remaining part is plain disc, so as to combine the
advantages of better impulse withstand at the high voltage end of
the winding and reasonable labour cost for the winding as a whole.
These are known as partially interleaved windings.
5.1.7 Rib Shielded Windings
An alternate way of increasing the series capacitance without actually

interleaving is achieved in Rib Shielded Windings. Floating shields
are provided inside continuous disc windings, and are comparatively
easier to manufacture, when compared to interleaved disc windings.
The shield wires are not conductively connected to circuit (Fig. 5.13).
S1
7 6 5 4 3 2 1
8 9 10 11 12 13 14
S2
21 20 19 18 17 16 15
22 23 24 25 26 27 28
S3
Figure 5.13 Rib shielded winding.
5.1.8 Shielded Layer Windings

This type of winding is generally used for star-connected transform-
ers having graded insulation and for voltages greater than 132 kV
class. The winding consists of a number of concentric spiral coils
arranged in layers. The layers are graded in lengths from the
longest at the neutral end (innermost layer) to the shortest at the
line end (outermost layer). The layers are arranged between two
concentric cylindrical shields, connected one to each end of the
winding (Fig. 5.14). All these layers are connected in series, wherein
two schemes are possible, viz. parallel-layer type (Fig. 5.14) and
tapered-layer type (Fig. 5.15).
The layers are separated by oil ducts and unbonded paper
cylinders. During winding, the latter are arranged to extend well
beyond the turns of the layer and afterwards these extensions are
petalled and bent over at right angles to form insulating flanges
between succeeding layers. These flanges provide an insulation
system to ground, which increases progressively from the neutral-
end to a maximum for the line-end of the coil.
The winding layers and the shields form a series of capacitors
and are so dimensioned that it results in substantially equal capaci-
tances in series. This ensures a relatively uniform distribution of
Neutral Line
shield shield
Figure 5.14 Shielded-layer winding (parallel-layer type).
Neutral Line
shield shield
Figure 5.15 Shielded-layer winding (tapered-layer type).
surge voltages throughout the winding. When the winding current

and the density of the leakage flux are not very high, the winding
can be wound with conductors of rectangular cross-section. For
large currents and higher density of leakage flux, transposed con-
ductors are used.
5.2 Surge Voltage Behaviour of Windings
In service, transformer windings are exposed to various transient

over-voltages like lightning surges, switching surges, etc. These
over-voltages have steep wavefronts and relatively longer tails. The
transformer windings respond to surge voltages as a system of ca-

pacitance and inductance network1. For the first few micro-seconds
after application of impulse, the winding behaves as a capacitance
network. Subsequently, the self and mutual-inductances of the
winding elements come into effect and impart oscillatory nature to
the voltage appearing at different parts of the winding. Thus it is
worth examining separately the initial voltage distribution due to
capacitive effect alone and the osciliatory behaviour due to combined
effect of capacitances and inductances.
5.2.1 Nature of Surge Voltages

A surge on a transmission system due to lightning discharge, either
in the vicinity of the line or to the line itself may have an extremely
complicated waveform. For the purpose of testing of transformers, a
standard waveform has been formulated for the full-wave impulse.
It has a wavefront of 1.2 ms and wavetail of 50 ms.
Impulse waves propagating along the line may occasionally create
flashover at the insulators to ground, causing sudden collapse of the
impulse voltage and the consequent very high rate of change of
voltage in the transformer windings. This type of wave is termed
chopped-wave impulse. The chopping of voltage occurs usually
between 2 to 6 ms.
Another kind of over-voltage is caused due to switching on and off
the loads or sources on the line and is known as switching surge.
With the advent of EHV systems operated at reduced BIL, the
switching surges are becoming an important factor.
The standard wave shape for the switching surges has a virtual
front time of at least 20 ms, a duration above 90% of specified
amplitude of at least 200 ms and a total duration to the first zero
passage of at least 500 ms.
The full-wave impulse and the switching surge waveforms can be
defined mathematically as the difference between two exponential
functions, e.g.
V(t) = a0 (e–b1t – e–b2t) (5.1)
the chopped wave has the same waveform up to the instant of
chopping and soon after that it reduces to zero. Such a waveform
can be defined as
V(t) = a0 (e–b1t – e–b2t) – b0 (e–b3t¢ – e–b4t¢) (5.2)
where, t¢ = t – tc
tc = instant of chopping
5.2.2 Idealization of Transformer Winding

In a winding there exists capacitance between the adjacent turns
within a disc or layer, capacitance between the adjacent discs or
layers, capacitance to ground and to other windings. Similarly, there
exists self- and mutual-inductances as pertaining to the individual
turns, the discs/sections, one part of the winding to another or
one whole winding to another. Although both the capacitance and
inductance are of distributed nature, for practical computation
purposes these have to be lumped in varying degrees according to
the desired accuracy. Also, the effect of the winding resistance is not
significant and is therefore neglected.
Figure 5.16 shows a part of a transformer winding represented as
a network consisting of
(i) series capacitances (Cs)
(ii) ground capacitances (Cg)
(iii) self-inductances (L)
(iv) mutual-inductances (M)
M1k M1n
L1 Lk Lk + 1 Ln
o M12 1 L2 2 K n
Csk + 1
Cs1 Cs2 Csk

ik
Cg1 Cg2 Cgk – 1 Cgk Cgk + 1 Cgn – 1 Cgn
Figure 5.16 Network representation of a winding.
5.2.3 Initial Voltage Distribution

On the incidence of the impulse wave, the inductive elements
behave like open-circuit elements and the winding can be treated as
a capacitive network as shown in Fig. 5.17. The node o is impulsed,
i.e. applied with voltage v(t) as in Eq. (5.1) or (5.2).
Let n = number of nodes of the network
Csk = series capacitance of the kth segment of the network
Cgk = ground capacitance of the kth segment of the network
ek = voltage appearing at the kth node
o 1 2 k–1 k k+1 n–1 n
Cs1 Cs2 Csk Csk + 1 Csn
V (t )
Cg1 Cg2 Cgk Cgn – 1 Cgn
Figure 5.17 Capacitive network representation of a winding

during initial voltage distribution.
v(t) = input voltage

t = time in microseconds
d
p = operator
dt
The node equation for node 1 can be written as
de de dv(t)
(Cg1 + Cs1 + Cs2) 1 – Cs2 2 = Cs1
dt dt dt
and for kth node as
dek dek + 1 dek - 1
(Cgk + Csk + Csk + 1) – Csk + 1 – Csk =0
dt dt dt
The equations for all the nodes can be expressed conveniently by
the matrix equation
LM e OP LMC
1 s1 . p . v(t) OP
e 0
C◊p M P = M PP
2
(5.3)
MM � PP MM �
PQ
Ne Q N
n 0
LM e OP
1 LMCs1 . p . v(t) OP
e 0
p M P =C MM PP
2 –1
or (5.4)
MM � PP MN �
PQ
nNe Q 0
where
LM(C g1 + Cs1 + Cs2 ) - Cs2 0...............0 OP
MM -Cs2 (C g 2 Cs2 Cs3 ) - Cs3 ........0
+ +
PP
�
C= M PP
MM �
MM 0 0......................................-Csn PP
N 0 0 - Csn - 1 (Cgn - 1 + Csn - 1 + Csn ) PQ
Equation (5.4) can be solved numerically (e.g. fourth order Runga

Kutta method).
For a winding having uniformly distributed series and ground
capacitance, i.e., Cg1 = Cg2 = ...... = Cgn = Cg and Cs1 = Cs2 = .... = Csn =
Cs it has been shown that the initial voltage distribution depends
on the value of a = C g / Cs . The voltage at a distance x from the
neutral end can be expressed as
sinh a ◊ x
V x = V m◊ (5.5)
sinh a
The initial voltage distribution for different values of a is shown
in Fig. 5.18 for windings with grounded neutral. For a = 0, the
distribution is perfectly linear. For other values of a the voltage
distribution is nonlinear and the voltage gradient at the line end is
a times the linear voltage gradient.
Voltage
a=0
a a=5
=
10
Line Neutral
end end
Figure 5.18 Initial voltage distribution along a winding

(Grounded neutral) for different values of a.
Due to this extra strain on the insulation between turns at the

line end of a high-voltage winding, sometimes the end-turns are
reinforced with extra insulating material. However, the end-turn
reinforcement is a matter of careful design because inappropriately
increased insulation thickness may result in undue increase of
impulse voltage at these turns and thereby defeat the purpose itself.
Use of electrostatic shields at the line-end of a disc winding helps
in improving the nonlinearity of the initial voltage distribution.
Sometimes these shields are used in the neutral end also. These
shields are connected electrically to the end-section. Being in close
proximity of the section, there exists a large capacitance between
the shield and the turns of the end-section. Due to this additional
capacitance, the initial voltage distribution in the end-section
becomes more linear.
For plain disc windings, a generally varies from 5 to 15. An outer
coil has less a compared to a similar inner coil as the latter faces two
ground planes (i.e. higher Cg).
The a of the windings can be reduced either by (a) decreasing the
Cg, or (b) by increasing the Cs. The latter approach is followed in
case of interleaved disc windings. Referring to Fig. 5.12 we find that
the voltage between two adjacent conductors is 2m. (V/T), where
m is the number of turns/section. The stored energy between the
adjacent conductors is
1
Q1 = C [2m (V/T)]2
2
1
= ◊ 4m2 C (V/T)2
2
Whereas the corresponding energy between the adjacent conductors
in a plain disc winding is
1
Q2 = C(V/T)2
2
Q1 = 4m2 Q2
In other words, the effective series capacitance of the interleaved
disc winding is 4m2 times more. This will bring down the a substan-
tially and result in nearly linear distribution.
The other approach, i.e. of decreasing the Cg is followed in
shielded layer windings. The portion of the winding seen by earth is
very small because of the concentric dispositions of the layers. Also,
equalization of the inter-layer (series) capacitance, by shortening
the length of layers corresponding to increased diameter, ensures
nearly uniform distribution of impulse voltage between the layers
(Fig. 5.19).
5.2.4 Transient Voltage Distribution

After the elapse of a few microseconds, when the rate of change of
voltage comes down considerably, the inductances come into effect,
in addition to the capacitances. The combined effect of inductances
and capacitances gives rise to oscillatory voltages at different parts
Ideal
Initial
Figure 5.19 Voltage distribution in shielded-layer windings.
of the windings.3,4 A multi-winding transformer can be represented

by a multiple-ladder network (Fig. 5.20).
3U1
2U1
1U1
Figure 5.20 Multiple-ladder network representation of a multi-

winding transformer.
It will be obvious that solution of such a network can be done only

with the use of computers. To illustrate the principles involved for
1.2
1.0
0.8
0.6
0.4
0.2
0
6 12 18 24 30 36 42
(a) Response near line end.
1.2
1.0
0.8
0.6
0.4
0.2
0
6 12 18 30 36 42
–0.2
–0.4
(b) Response at the middle of the winding.
0.4
0.2
0
6 12 18 42
–0.2
–0.4
(c) Response near the neutral end.
Figure 5.21 Computer-aided analysis of impulse voltage

distribution in transformer windings.
computation, we take a simple network as shown in Fig. 5.16. The

analysis presented here is in line with that of Hartill et al5. The
network parameters are given as:
LM L 1 M 12 M 13 ......... M 1n OP
M M 21 L2 M 23 ......... M2 n PP
P =M �
(inductance M
matrix) MM � PP
NM n1 M n2 M n3 .......... Ln PQ
where Li = Self-inductance
Mjk = mutual inductance
LM 1 0 0....... 0OP
MM Cs0 1
1 0........ �PP
S =M Cs2 �P
(series capacitance MM � �P
matrix) 1 P
MM � ........
Cs Pn PQ
N0 0
LM F 1 + 1 I -
1
0 ...
OP... 0
MM H G C s1 C J
K
g1 C g1 PP
1 F 1 + 1 + 1I 1
Q = MM -
C GH C C C JK -
C
... 0 P
PP
g1 g1 s2 g2 g2
(combined MM �
series and �
ground MM � �
PP
capacitance
N.................... .............................. ...................PQ
matrix)
Let ik = lower mesh current in kth mesh

i¢k = upper mesh current in kth mesh
fk = Ú ik dt
f k¢ = Ú i¢k dt
The solution can be written as

p2f = (Gp2 + H) V(t) + F (5.6)
d2
where p2 =
dt 2
G = Q–1
H = Q–1 SP–1
F = Q–1SP–1 (S – Q)
Equation (5.6) can be solved numerically (say by 4th order Runge-
Kutta Method)6. Figure 5.21 shows the responses at various points
in a transformer winding under impulse voltage, as obtained by this
method.
5.3 Internal Heat Transfer in Windings
The heat generated in the transformer winding is transferred to oil

mainly by convection. The oil which is in contact with windings
takes the heat from the latter and becomes warmer and lighter.
The oil, now lighter, rises upward and finally goes to the cooling
equipment, where it gets cooled. Colder oil enters the coils from the
bottom of the windings and this way, continuous circulation goes on.
The heat from the inner parts of the winding is transferred to its
outer parts by conduction. The thermal conductivity of copper is
many times higher than that of the insulating paper on the conduc-
tor. It is thus obvious that paper tries to impede the conduction of
heat and the thicker the paper, the more is this impedance. The
thickness of paper covering on the conductor has to be a balance
between the requirement of electrical insulation and the thermal
impedance.
The cooling ducts within the windings provide a path for oil circu-
lation. The ducts are either axial (Fig. 5.22) or radial (Fig. 5.23).
The typical paths of oil flow are also shown therein. If the radial
dimension of the coil is very large, it is advantageous to introduce
an intermediate vertical cooling duct (Fig. 5.24). The effectiveness
of these intermediate ducts can be increased by staggering their
location in alternate sections.
The heat transfer can be substantially improved if the oil flow
is properly directed within the coils. Figure 5.25 shows a typical
arrangement wherein baffles have been provided after a few sections
Axial
duct
Figure 5.22 Axial oil flow duct. Figure 5.23 Radial oil flow duct.
Intermediate
cooling duct
Figure 5.24 Intermediate vertical duct improves cooling.
to direct the flow of oil in a unidirectional manner between two

baffles. Oil pumps are used externally to create increased flow of oil
within the windings. Oil with increased and unidirectional flow can
take away heat more effectively from the windings. This results in
lower temperature gradient for the windings.
5.4 Insulation Design
For oil immersed transformers, the insulation system comprises

a mixed dielectric, viz. oil and cellulosic material. The insulation
Baffles
Figure 5.25 Directed oil flow improves cooling.
structure can be broadly categorized as per their location in trans-

former as follows:
(a) Minor insulation refers to insulations between different parts
of one winding, like insulation between turns, layers, etc.
(i) The insulation of the conductors is generally of paper,
which is wrapped around the conductor. For continuously
transposed conductor, the individual strands are coated
with a layer of enamel before final wrapping of paper.
(ii) Insulation between turns is provided either by conductor
insulation (e.g. in spiral windings, in turns within a section
of disc winding) or by conductor insulation along with
radial oil ducts formed by using block (e.g. for helical
coils, or for insulation between sections of a disc winding).
The thickness of the insulation is determined by the
voltages (power frequency as well as impulse) appearing
between different conductors, whereas the thickness of
the oil duct is determined from voltage as well as thermal
considerations.
(iii) Insulation between the layers of the shielded layer winding
comprises vertical oil ducts and paper cylinders.
(iv) In high voltage windings, the end turns are sometimes

reinforced to take care of nonlinear distribution of impulse
voltage (see Sec. 5.2.3).
(b) Major insulation comprises insulation of windings to earth
and transformer core, other windings of the same phase (e.g.
HV winding to LV winding) and between one phase and
another. The insulation between different windings and
inner winding to core consists of pressboard cylinders
separated by oil ducts. Pressboard barriers are provided
between windings of different phases and between the
windings and the tank. Ends of windings are insulated from
yokes by adequate number of angle rings/angle washers
depending upon the voltage class. For shielded layer
windings, the flanges obtained by petalling the radial paper
serve as the angle rings.
5.4.1 Composite Dielectric Insulation System

Some important considerations regarding design of composite di-
electric insulation system are discussed below:
(a) Oil
(i) In a composite dielectric system, the dielectric with lower
permittivity bears more than average voltage stress. The
permittivity of transformer oil is nearly half that of
pressboard. Therefore, the electric stress is nearly twice as
great in the oil in the annular ducts as on the pressboard
cylinders for the same thickness of insulation.
(ii) The electric strength of oil is substantially less than that of
cellulosic material.
(iii) Oil ducts show an important characteristic of voltage
withstand, that narrower the duct, higher is the stress
withstand level (kV/mm).
Under the influence of electric field, foreign substances
in oil in the form of dust, moisture, etc., have a tendency to
align themselves in radial lines, giving rise to paths of low
dielectric strength, with consequent danger of breakdown.
(iv) Due to the characteristic mentioned in (iii) above, the

structural stability of the barriers, which form the oil ducts,
is of great importance.
(b) Solid Insulation

(i) The solid insulation is oil impregnated under vacuum. There
is a considerable difference in the dielectric strength of
pressboard impregnated under a relatively high and a
relatively poor vacuum. Thus a high vacuum is desirable
during impregnation.
(ii) Proper placement of the insulation is the foundation of
successful insulation structure. It is desirable that
insulating materials are subjected only to breakdown stress,
and that high creep stresses on the boundary layers of the
barriers are avoided. Therefore, the boundary layers of solid
insulation should correspond to the equipotential planes as
far as possible. Since, however, it is not possible to entirely
avoid the creep stresses, the designer has to limit the
stresses within permissible values.
(iii) Due to imperfections in their manufacture, the pressboard
sheets may contain some micro-voids, which tend to lower
the withstand strength. Since the probability of locations of
such voids coinciding is very less, it is better to built up a
required thickness of solid insulation from more than one
sheet.
(iv) Insulating materials exhibit some degree of dielectric losses
when placed in an electric field. The dielectric losses are
dependent on the voltage and the frequency of the field and
the dielectric constant and the loss angle of the material. The
heat generated due to these losses need to be dissipated,
otherwise it may lead to undue temperature rises.
(c) Partial Discharges

Partial discharges are regarded as the most hazardous phenomenon
with respect to the service life of the insulation. Two types of partial
discharges may occur in the solid insulations.7
(i) Partial discharges of high intensities may take place on the
surface of the pressboards. However, these are considered
less hazardous, as the breakdown time is substantially

longer.
(ii) Concentrated partial discharges may occur on the sharp
edges of electrodes. These discharges gradually penetrate
deeper and deeper into the pressboard and eventually lead
to a breakdown.
The large oil ducts may also give rise to partial discharges when
the stresses appearing in it exceed the limits of partial discharge
inception voltage.
A sound design practice would be to aim at completely partial
discharge free insulation structure for a given over-voltage.
5.5 Electric Field Plotting
From the above, it is clear that a sound design of an insulation sys-

tem requires pre-determination of the voltage stress levels (both
creep and puncture) under impulse and power frequency voltage
conditions. The voltage stresses are a function of the shapes and
relative dispositions of the windings, shapes of insulation structure
and the dielectric media. The voltage stresses can be computed from
the knowledge of electrostatic field strength at various points
within the structure. Electrostatic field is Laplacian in nature8 and
the different methods of solution are:
5.5.1 Analytical Method

Where the electrode and insulation configuration is simple, a
closed-form solution of the field equation can be obtained directly.
For certain more complex cases, closed-form solutions can still
be obtained by techniques like imaging, conjugate functions,
conformal transformation, etc.
5.5.2 Graphical Method
In this method, successive graphical approximations are made on

the orthogonal properties of flux and equipotential lines. However,
this method becomes too tedious for any problem of practical signifi-
cance.
5.5.3 Analogue Methods

Various analogues have been used by different workers, like, resis-
tance network, rubber membrane, sand models, resistance paper
and electrolytic tank. The latter two are the most well known. Each
of these has its own limitations, with the electrolytic tank method
being the most versatile one. But the need to make expensive and
time consuming models for each problem and also the problems like
polarization, surface tension, etc., seriously limit even this method.
5.5.4 Numerical Methods

In a practical problem, as in transformer, with its inherent complexi-
ties, like electrodes with odd profiles, as multitude of intervening
dielectrics, the analytical methods become inadequate and this
leads us to numerical methods of solution. The latter are always
approximations to the true solutions, but with sufficient care yield
results that are true within engineering accuracy.9–11 The two
main numerical methods are the finite-difference method and the
finite-element method. A brief description of the finite-difference
formulation of electrostatic field problem is given below.
5.6 Finite Difference Method (FDM)
In this method, the continuum is replaced by a mesh system of

discrete points. Consider Fig. 5.26 which shows a general point
A0 of the mesh system, surrounded by the four points A1, A2, A3 and
A4. The arm lengths are p, q, r and s as shown. Each mesh has
two dielectric media interfacing at the diagonal, therefore a total
of eight dielectrics meet at the point A0. Consider a box whose
cross section is the contour 1–2–3–4–5–6–7–8, passing through the
midpoints of the arms, and with unit length in the direction into the
paper.
The Laplace’s equation, —2f = 0, in the charge-free region
becomes
ÚÚ D ◊ n ◊ dl = 0 (5.7)
Since D = e E and that unit length into paper has been taken,
Eq. (5.7) can be written as
f e E dl = 0
1- 8
A2
e3 e2
4 3 2
q
e4 e1
A3 5 1 A1
A0
e5 e8
8 s
6 7
e6 e7
A4
r p
Figure 5.26 A general point A O in the mesh system.
This integral consists of 8 segments, e.g. 1–2, 2–3, etc., up to 8–1.

Writing this integral equation in the finite difference form and
noting that E = – grad, A, we get
( A0 - A1 ) e 8 s e 1q
LM OP( A - A2 ) e 2 p e 3 r LM OP
+ + 0 +
p 2 N 2 Q q 2 2 N Q
( A0 - A ) Le q e sO (A
3 4 5 0 - A ) Le r e pO
4 6 7
+
r MN 2 + 2 PQ + s MN 2 + 2 PQ = 0 (5.8)
Putting C1 = [e8s + e1q]/2p

C2 = [e2p + e3r]/2q
C3 = [e4q + e5s]/2r
C4 = [e6r + e7p]/2s
The Eq. (5.8) can be written as
A0 (C1 + C2 + C3 + C4) = A1C1 + A2C2 + A3C3 + A4C4
4
ÂAC i i
i=1
or A0 = 4
ÂC i
i=1
where A = scalar electrostatic potential function

D = electric flux density
E = electric field intensity
e = permittivity of dielectric medium
This equation for the potential A0 is successively applied to every
node in an iterative manner, till sufficient accuracy is obtained. It is
to be noted that this equation is true in charge-free regions only.
Thus, whenever a node which is part of a conductor with definite
potential, is encountered in the iterations, the computation is
skipped for that node.
The electrostatic field problem is essentially a boundary value
problem and the nature of the boundaries are to be known apriori.
The two types of boundaries met are Dirichlet and Neumann bound-
aries. For the former, the potential on the boundary is a constant,
whereas the Neumann boundary is characterized by constant
gradient condition. Any axis of lateral symmetry in the problem
region can be thought of as a Neumann boundary.
5.6.1 Use of Digital Computers

A generalized computer program based on the above algorithm can
cater for almost all types of insulation structures. The grid lines are
so chosen that all interfaces between dielectrics and conductors,
etc., coincide with grid-lines. The curved lines can be approximated
by segmenting them into a number of straight lines. The grid
system can be closely spaced where higher accuracy is desired.
The problem is defined through a set of input data as follows:
(a) Number of horizontal and vertical grid-lines and their
dis-tances
(b) Position of the insulation and the conductors
(c) Permittivities of the media.
5.6.2 Special Features

Since the formulation is in terms of potentials at discrete points
within the region, it is possible to consider both, point conductors
and conducting foils exactly as they occur in the problem. This
is of particular importance in the case of transformers where
electrostatic shields, which are essentially foil-conductors, are used
to a large extent. A special feature incorporated in the program is
the computation of potential acquired by a floating conductor in the
region of the field.12
5.6.3 Application of Field Plotting Technique

to Transformer Insulation Design
To illustrate the nature of the fields encountered in transformers, a
typical field plot of the coil-end region of a high voltage transformer
is shown in Fig. 5.27. The different coils have been shown along with
the equipotential lines in steps of 10%. According to the nature of
the field, three zones have been marked as A, B and C.
The zone A, or the vertical zone has a nearly uniform field.
The number and thickness of the barriers is decided by the stress-
withstand of the oil gaps. The zone B has nonuniform field and is
generally of horizontal configuration. In this zone the oil gaps are
larger than in zone A. The placement of barriers and static rings
tend to make the field more uniform.
LV
A HV R
Figure 5.27 Typical field plot of the coil end region in a high
voltage transformer.
The zone C is the transition zone between zones A and B and is

the most important from insulation design consideration. It is a
zone of high gradients and is subject to both puncture and surface
creep stresses. It is relatively easy to guard against puncture,
but minimization of surface creep stresses requires the knowledge
of the exact positions of the equipotential lines. If the insulation
profiles are designed to conform to the shape of equipotential lines,
creep stresses are minimized.
Figure 5.28 shows another example of electrostatic field plotting
for an EHV transformer.
10%
20%
30%
40%
A B 50%
Angle ring
Angle ring 60%
Angle ring
70%
80%
Angle ring
650 KVP 85.5%

650
650 KVP
KVP
650 KVP 550 KVP
72.4%
100%
KVP
650 KVP
760
(85.5%)
Coil No. 1
Coil No. 2
Coil No. 3
Coil No. 4
Coil No. 5
10%
20%
30%
40%
50%
60%
70%
80%
90%
Figure 5.28 Electrostatic field plot inside transformer top end of

coils under impulse test.
Once the knowledge of the field is obtained, it is necessary to

ensure that the stresses appearing at various parts of the insulation
structure are contained below the limits of partial discharge
inception.13 For achieving this, the designer resorts to any one or all
of the following methods to bring the stresses within limits.
(a) Electrodes having very sharp corners produce undesirably
high stresses around the corners. It is therefore imperative
to avoid such sharpness in electrodes. The minimum
allowable radius of electrodes is governed by their relative
disposition.
(b) Though it may suffice from thermal considerations to use
a small size conductor for low current, high-voltage leads, it
may sometimes be necessary to go in for a conductor of a
larger diameter, as the latter gives rise to lower electric
stresses.
(c) In oil immersed electrode systems, where the stress is
considerably high on the surface of the electrode and falls off
exponentially, the electrode is covered with solid insulation

to take advantage of its higher withstand capability and to
push the oil insulant to a safe point where it is subjected to
allowable stress limits.
(d) Should the above technique be not viable, withstand
capacity of oil is considerably increased without lowering the
stress to which it is subjected to, by dividing the highly
stressed oil zone into a number of thin laminae by using
barriers of solid insulation.
(e) In some important areas, use is made of moulded insulation
items whose contours are made in accordance with the shape
of the equipotential lines. Thus surface creep stress can
almost be avoided and a very compact insulation system can
be evolved.
(f) Another source of partial discharge inside the transformer is
the cables and the terminal gear. Here also, it is necessary to
ensure that the electrical stresses at the copper-paper
interface or paper-oil interface are within limits. Provision of
too much of insulation on the copper, however, impedes the
heat transfer.
REFERENCES
1. Heller and Veverka, Surge Phenomenon in Electrical Machines,
Ileffe Books Ltd, London.
2. Blume, L.F. et al, “Transformer Engineering”, John Wiley & Sons,
New York, 1965.
3. Lovas-Nagy et al., “A Matrix Method of Calculating the Distribution
of Transient Voltages in Transformer Winding”, Proc IEE, Vol. 110,
pp 1663, 1963.
4. Miki, Hosoya and Okuyama; “A Calculation Method for Impulse
Voltage Distribution and Transferred Voltage in Transformer
Windings”, Trans IEEE, PAS-97, No. 3, May/June 1978.
5. Dent, B.M., Hartil et al.; “Method of Analysis of Transformer
Impulse Voltage Distribution Using a Digital Computer”, Proc IEE,
Vol. 105, Part A, 1958.
6. Smith, G.D.; “Numerical Solution of Partial Differential Equations:
Finite Difference Methods”, Oxford Press.
7. Moser, H.P., “Transformer board”, Scientia Electrica.
8. Binns, K.J. and Lawrenson, P.I, “Analysis and Computation of

Electric and Magnetic Field Problems”, Pergamon Press, 1973,
9. Galloway, R.H., Ryan, H.M. and Scott, M.F., “Calculation of Electric
Fields by Digital Computers”, Proc IEE, Vol. 114, pp 824–829, 1967.
10. Ryan, H.M., Mattingley, J.M. and Scott, M.F., “Computation of
Electric Field Distribution in High Voltage Equipment”, Trans
IEEE, PAS, pp 148–154, 1971.
11. Cermak, I.A. and Silvester, P. “Boundary-Relaxation Analysis
of Rotationally Symmetric Electric Field Problems”, Trans IEEE,
PAS-89 (5/6), pp 925–932, 1970.
12. Mittal, M.L., Sarkar, S.C. and Kumar, P.R., Computerised
Electric Stress Analysis of Insulation Structures in Transformers”,
International workship on EHV and UHV insulation (Instac-82),
Institution of Engrs (India) Ltd., 1982.
13. Mittal, M.L. and Sarkar, S.C., “Recent Design Practices in Power
Transformers”, International conference on transformers, IEMA,
New Delhi, 1982.
C HAPTER 6
Voltage Regulation
and Tapchanger
B. L. Rawat
A.K. Ekka
Voltage variation in electrical systems is a normal phenomenon,

because of rapid growth of industries and distribution network. It
is very essential to maintain the system voltage within prescribed
limits for the better health of electrical equipments. Voltage of the
system can be varied by changing the turn ratio of transformer. The
device tapchanger is used for adding or cutting out turns of primary
or secondary winding of the transformer. Basically tapchanging
equipment can be divided in two categories:
(a) Off-circuit tapchanger,
(b) On-load tapchanger.
6.1 Off-Circuit Tapchanger
The cheapest method of changing the turn ratio of a transformer is

the use of off-circuit tapchanger. As the name implies, it is essential
to de-energize the transformer before changing the tap.
An off-circuit tapchanger, as shown in Fig. 6.1, consists of princi-
pally the following three parts:
(a) Operating handle projecting outside the transformer
(b) Fixed contact with connecting terminal
(c) Insulating shaft with moving contact system
The basic transformer winding circuit arrangements using
Voltage Regulation and Tapchanger 139
Manual
operating
mechanism
Insulating
shaft
Fixed
contact
Moving
contact
Figure 6.1 Off-circuit tapchanger.
off-circuit tapchanger are as shown in Fig. 6.2. They are:

(a) Linear
(b) Single-bridging
(c) Double-bridging
(d) Series-parallel
(e) Star-delta
Depending upon the requirement, any of the above arrangements
of the winding can be made use of to get desired voltage regulation.
To prevent unauthorized operation of an off-circuit tapchanger, a
mechanical lock is provided. Also to prevent inadvertent operation,
6
5
4
3
5 2 2
4 3 1
3
2 4 5
2 4
1
3 1
1 5 X
X 6
X
(a) Linear (b) Single-bridging
4
8
12 4
11 10 11
10
9
8 9 2
7 3
7
12 3
2
6 2
5
4 4 5
3
2 3
1
3 1
X 1
X 6
(c) Double-bridging (d) Series-parallel
X
X 2
2 1
(e) Star-delta
Figure 6.2 Off-circuit tapchanger connection circuits.

an electromagnetic latching device or microswitch is provided to

open the circuit breakers to de-energize the transformer while
operating the handle of tapchanger before movement of contacts on
tap switch.
6.2 On-load Tapchanger (OLTC)
On-load tapchangers are employed to change turn ratio of trans-

former to regulate system voltage while the transformer is delivering
normal load. With the introduction of on-load tapchanger, the
operating efficiency of electrical system has considerably improved.
Now-a-days, almost all the large power transformers are fitted with
on-load tapchanger.
All forms of on-load tapchanging circuit possess an impedance,
which is introduced to prevent short circuiting of tapping section
during tapchanger operation. The impedance can be either a resistor
or centre-tapped reactor. The on-load tapchanger can in general, be
classified as resistor or reactor type.
6.2.1 Reactor Transition Type OLTC
In early designs, the use of centre-tapped reactor as tapchanging

impedance, was in general more popular, inspite of the inevitable
shorter contact life. One of the principal advantages of mid-point
reactor transition is that, twice as many active working positions
as that of transformer tappings could be obtained. This can be of
considerable advantage where large number of tapping positions are
required. Reactor transition type on-load tapchangers are manufac-
tured and used only in the United States of America. In other parts
of the world, resistor transition type on-load tapchangers are being
manufactured and used.
6.2.2 Resistor Transition Type OLTC
Resistor transition has considerable advantage of longer contact

life, due to relatively short arcing time associated with unity
power factor switching. With the introduction of high speed resistor
transition tapchanging, it is possible to break the arc at first to
current zero. Furthermore, the introduction of contacts using
copper/tungsten alloy arcing tips has brought about a substantial

improvement in contact life.
Transition resistor type tapchangers can be divided into two
types, those which carry out selection and switching on the same
contacts, and those which have tap selectors and a separate diverter
switch. The first category is known as single-compartment type,
while the second one as double-compartment type.
6.2.3 Single-compartment Type OLTC
The single-compartment type employs a rotary form of selector

switch with single-transition resistor or double-transition resistor.
Figure 6.3 illustrates the switching sequence in moving from one
tap to the next tap employing single-transition resistor. This
switching cycle is known as “asymmetrical pennant cycle”.
Tapchangers constructed with one transition resistor are suitable
for power flow in one direction only. This particular contact
arrangement is not suitable for power flow in reverse direction.
Figure 6.4 illustrates the switching sequence in moving from one
tap to the next tap employing two transition resistors. This switch-
ing cycle is known as Flag cycle and the arrangement is suitable
for bi-directional power flow. Single compartment tapchangers
available presently are suitable for currents up to 600 A and 66 kV
voltage class (with certain limitation up to 132 kV class) winding of
transformers.
6.2.4 Double Compartment Type OLTC
On larger transformers, the on-load tapchanging equipment is more

usually arranged with separate tap selectors and divertor switches.
The tap selectors are generally arranged in a circular form. The
divertor switches have contacts operating in rapid sequence with
usually four separate make and break units.
Figure 6.5 shows typical selector and switching arrangements
with on-load tapchanger on neutral end of the star-connected
winding of a transformer.
Figure 6.5(a) shows linear selector arrangement for 16 steps
and 17 positions. In the diagram, the on-load tapchanger is shown
connected to tapping 4. While changing from tapping 4 to tapping 5,
M1 is opened first and this transfers the load current via A1 with
3
2 4
(a) Tap position No. 2. The main contact is
1
carrying the load current. The transition

contact is open and rests between the
fixed contacts 2 and 3.
3
2 4
(b) Transition contact makes on the
1
fixed contact 3, the transition resistor

bridges 2 and 3 and carries
circulating current.
3
2 4
(c) The main switching contact breaks

1
and the transition resistor carries

the load current.
3
2 4
1
(d) The main switching contact makes on

contact 3 and carries the load current.
3
2 4
1
(e) The transition contact opens. The

tapchange operation complete.
Figure 6.3 Switching sequence.

3
2 4
(a) Tap position No. 2. The main contact is

1
carrying the load current. The resistor

contacts M1 & M2 are open, resting
M2
M1 between fixed contacts.
3
2 4
(b) The resistors contact M1 has made
1
on the fixed contact 2 and the main

contact has broken the transition
resistor contact M1 carries load
M1 M2
current.
3
2 4
(c) The resistor contact M2 has made on
1
the fixed contact 3. The load current

is divided between resistor contacts
M1 and M2. The circulating current is
M1 M2
limited by resistors.
3
2 4
(d) The resistor contact M1 has

1
broken from the fixed contact 2.

The transition resistor contact
M1 M2 M2 carries the load current.
3
2 4
(e) Top position No. 3. The main contact
1
has made on fixed contact 3. The

resistor contact M2 has broken
M1 M2 from the fixed contact 3. The main
contact is carrying load current.
Figure 6.4 Switching sequence.

8 steps
16 steps
8 steps
8 steps
R1 R2 R1 R2 R1 R2
M1 A1 A2 M2 M1 A1 A2 M2 M1 A1 A2 M2
Voltage Regulation and Tapchanger
N N N
Diagram a Diagram b Diagram c

145
Figure 6.5 Linear coarse-fine and reversing switch type connection diagram.
resistor R1 in series. Then A2 closes and two resistors R1 and R2 are

in series across tappings 4 and 5. A circulating current will flow
through these resistors because of step voltage between tappings
4 and 5. The load current is divided passing through each of the
resistors to each of the tappings. A1 then opens and interrupts the
circulating current and the load current is transferred to tapping 5,
passing through resistor R2.
Finally M2 closes and takes load current, completing the
tapchange. This sequence of tapchange in most of the designs takes
40 to 80 ms. For a tapchange in the opposite direction the sequence
is reversed. For successive tapchange in the same direction, first the
tap selector contacts movement takes place. However, for the first
reversal, the tap selector does not move. This feature is obtained by
using a lost motion coupling in the mechanical drive of tap selector.
6.2.5 OLTC with Change-over Selector
When the tapping range is large it is advantageous to halve the

length of tapping winding and to introduce a reversing or change-
over selector. In Fig. 6.5(b) the tapped portion of the winding is
shown divided into eight sections and a further untapped portion
has a length equal to eight sections. In Fig. 6.5(c), the tapped section
of the winding itself is reversed to get double steps. Decision regard-
ing use of on-load tapchanger to adopt am of the two circuit designs
will depend on the design of the transformer.
With tappings on the neutral end of HV winding of three-phase
transformer, only one three-pole tapchanger is required per trans-
former. However, for large regulating transformers, single pole
tapchanger for individual phase shall be required. Three such
tapchangers are mechanically coupled together and driven by a
common driving gear.
6.2.6 Methods of Potential Connections
Figure 6.6 shows a typical selector and switching arrangement

generally followed by transformer design engineers depending upon
transformer specification requirements. During change-over opera-
tion of reversing switch or coarse tap selector, the tapping winding
gets temporarily disconnected from the main winding. During this
period tap winding acquire a potential determined by the winding
capacitances. The difference in voltage will, therefore, appear

during contact separation of the change-over selector as a recovery
voltage. If the calculation results show recovery voltage more than
the specified limit declared by the tapchanger manufacturer for the
particular design, the tapping winding should be connected continu-
ously or during changeover operation to a fixed potential.
This is possible by the following methods shown in Fig. 6.6.
RP
RP SP
N N N
(a) With installed (c) With tie-in resistor RP (c) With double
tie-in resistor RP and potential switch SP reversing switch
Figure 6.6 Method of potential connection.
(a) Potential connection by a continuously connected ohmic

resistor (tie-in resistor).
(b) Potential connection by a ohmic resistor which will be
inserted during changeover operation by means of a potential
switch.
(c) Double reversing switch (available with a particular special
design only).
6.2.7 OLTC with Coarse-fine and Reversing Arrangement
Comparing reversing with coarse-fine arrangement, the latter gives

lower copper-losses at the minimum number of turns position.
However, coarse-fine arrangement requires a more sophisticated
winding layout from dielectric point of view. Further, the wide
variation of impedance over the tapping range can be avoided by the

use of coarse-fine arrangement.
6.2.8 OLTC for Delta-Connected Winding
If the tappings are to be provided on the delta-connected winding,

usually three isolated single-pole design tapchangers are employed.
The connection diagram is shown in Fig. 6.7. Special arrangement of
tappings shown in diagram (c) can employ one two-pole tapchanger
and one single-pole tapchanger, mechanically coupled and operated
by single driving gear. From economic consideration, it is preferable
to employ tapchanger on the neutral end of star-connected winding.
Presently, tapchanger employed on delta-connected winding are of
132 kV class or less.
(a) Three-pole mid-winding (b) Three-pole line-end (c) Two-pole and one-pole
arrangement arrangement line-end arrangement
Figure 6.7 Tapping arrangement on delta-connected winding.
6.3 Constructional and Operational Features of OLTC
As described earlier, the on-load tapchangers mainly fall in two

categories:
(a) Single-compartment design
(b) Double-compartment design
6.3.1 Single-Compartment Design
In single-compartment design, same contacts are employed for

selection and transfer of current. All the contacts are housed in the
same compartment. The insulating oil of this compartment is kept

isolated from the main transformer oil. Two designs are available in
single-compartment type on-load tapchangers.
In one design the electrical contacts, transition resistors and
mechanical gear arrangement are housed in a pressure tight insula-
tion cylinder which is housed alongwith the transformer in the
same tank. Motor drive for manual/electrical operation is mounted
on the tank outside.
In the other design, electrical contacts, transition resistors and
geneva gears are housed in a steel tank, which in turn is mounted
on the side of transformer tank. The connection between transformer
tapping leads and tapchanger is through terminal board of epoxy
resin moulding. This terminal board is mounted either on trans-
former tank or tapchanger housing, depending on the design and it
also acts as a barrier to separate the insulating oil of transformer
and tapchanger.
6.3.2 Double-compartment Design
Large transformers of high voltage class employ double-compart-

ment type on-load tapchangers. The compartment which houses
make-break contacts and transition resistors is called diverter
switch compartment [as shown in Fig. 6.8 (Plate 3)]. The tap selector
which is the other part of the tapchanger is housed in a second
compartment. The modern design based on Jenson principle
presently available uses insulating pressure-tight cylinder to house
make-break switches, transition resistors and energy accumulators.
The tap selector with geneva wheel gear is without housing. The
diverter switch and tap selector are mechanically coupled and
housed alongwith transformer in the same tank. The tap selector
contacts for each phase are arranged in two circles, one having even
numbered contacts and the other odd numbered contacts. To achieve
alternating movement of tap selector contacts, geneva wheel gear
train (Fig. 6.9) is employed.
As the tap selector is housed in the same tank along with trans-
former, the Buchholz relay of transformer takes care of tap selector
also. However, for diverter switch a separate oil surge relay is
provided because the oil in diverter switch compartment is not in
contact with transformer tank oil. This oil surge relay trips the
Figure 6.9 Geneva wheel gear train.
transformer for any electrical fault that takes place in the diverter
switch compartment.
6.4 Manual and Electrical Operation of Tapchanger
A motor drive unit for operation of on-load tapchanger is supplied

and fitted on the side of the transformer tank. Motor drive housing
is designed for outdoor duty. The motor drive housing contains
all mechanical and electrical parts necessary for operating the
tapchanger [as shown in Fig. 6.10 (Plate 3)].
To facilitate manual operation of tapchanger, a hand crank kept
in the motor drive unit is to be pushed on the shaft so that it engages
the gearing. While pushing the crank on the shaft, safety switch
operates and breaks and supply to motor and the control circuit.
After the tapchanger reaches end position, further manual operation
results in the interruption, disengaging the drive shaft after
about two to three revolutions of the crank handle. The gears get
re-engaged after operation of crank handle in the opposite direction.
To electrical control of motor drive follows the step-by-step
principle, i.e. after energization of electrical control and starting of
driving motor, the tap change operation is accomplished independent
of whether the push button or switch have been operated during the
running time of motor drive. Another switching operation is possible
only when the control system is again in the rest position. Shortly
before the motor drive arrives in the end position, the limit switch
opens so that motor and its control can no longer be energized.
6.5 Automatic Control of Tapchanger
Automation requires the use of automatic voltage control at substa-

tion, so that a predetermined constant or compensated busbar
voltage can be maintained. In general, a tapchanger is provided on a
transformer for maintaining a predetermined outgoing voltage
where the incoming voltage is subjected to variation due to voltage
drops and other system variations. The output of the voltage
transformer connected to controlled voltage side of the transformer,
is used to energize automatic voltage regulating relay.
The voltage regulating relay does not issue a control signal, as
long as the voltage to be controlled remains within preset limits.
Once, however, the voltage deviation exceeds the set limit, a control
signal RAISE or LOWER is given to initiate tapchanger operation.
It is usual to introduce a time delay element within the relay itself,
to prevent unnecessary operation or hunting of the tapchanger
during transient voltage change. As an additional feature, load
compensation is frequently added. The load drop compensator
serves for compensating the resistive and the reactive voltage drops
of an outgoing feeder. This device keeps the voltage on-line end
constant, regardless of the load (amplitude and phase angle of
current). One tapped resistor and a tapped reactor, provided in the
load compensator of voltage regulating relay are fed from the
secondary of a current transformer. The primary winding of this
current transformer carries the load current and is arranged to
subtract a voltage drop proportional to the load on main transformer
from voltage applied to the voltage regulating relay. By suitable
adjustment of the resistance and reactance components, which will
depend upon the outgoing line characteristics, it is possible to obtain
constant voltage at some distant point on a system, irrespective
of the load or power factor. Figure 6.11 shows the principle of the
regulating relay with compensator.
6.6 Tapchanger Selection
While selecting a tapchanger for a particular transformer, following

points are to be considered:
— Voltage class of transformer winding and its rating.
Transformer
To load
C.T.
V.T.
Potentio- Reactor
meter
Voltage coil of regulating relay
Figure 6.11 Automatic voltage regulating relay with line drop

compensation-connection diagram.
— Percentage voltage variation required.

— Maximum through current.
— Step voltage between adjacent contacts.
— The switching capacity (maximum through current, step
voltage).
— Insulation level to ground and between various contacts.
— Number of steps and basic connection (linear, reversing or
coarse-fine).
— Temporary overloads.
— Short-circuit strength required.
— Number of operations required (any special duty).
6.7 Latest Trends in Tapchanger Design
At present tapchangers are available for the highest insulation

level of 1475 kVp impulse and 630 kV power frequency voltage.
Efforts are being made to develop tapchangers suitable for still
higher insulation level class. Further efforts are being made for
developing tapchangers smaller in size and having high reliability
and performance. The use of vacuum switches in the diverter
switch of tapchanger is being tried to increase its performance.
Also, thyristorized tapchanger will be available in future for very
special applications where excessively high number of operations
are required.
REFERENCE
1. Stigant S. Austeen and A.C. Franklin, The J & P Transformer Book,
Newness–Butterworth, London.
C HAPTER 7
Electromagnetic Forces
in Power Transformers
M. V. Prabhakar
T.K. Ganguli
The most severe mechanical stressing occurs in a transformer,

when it is subjected to a sudden short circuit. Since the currents
flowing through the windings at that time are enormous, the forces
generated are also enormous. Much work has been done to analyze
and calculate these forces. The problem does not yield easy solution
due to the transient and dynamic nature of the phenomenon.
7.1 Leakage Flux in a Typical Two Winding Transformer
The typical leakage flux pattern in a two winding transformer of

core-type is given in Fig. 7.1. The flux lines are almost axial at
the middle of the winding and start bending in the radial direction
as we move towards the winding ends. In other words, the axial
component is predominant along the height of the winding, except
at winding ends (top and bottom), where the radial component
becomes significant. The direction of current in the winding conduc-
tors is perpendicular to the flux lines, thus resulting in the axial and
radial component of forces.
7.2 Nature of Forces
Forces are produced due to the interaction of the current and the
magnetic flux density vectors. Thus, in general, the force vector can
have any direction. Nevertheless, it is easier to speak of the radial
Electromagnetic Forces in Power Transformers 155
Secondary coil
Primary coil
Next limb/tank
Core
*
N1 N2
Figure 7.1 Typical leakage flux pattern in a two winding transformer.
forces and the axial forces in a transformer, since these two compo-
nents of the force can be calculated and analyzed independently.
Also, the two components have influence on different parts of the
total transformer and it is necessary to obtain the two components
for design purposes. The nomenclature axial and radial is applicable
to concentric wound core type transformer and is assumed in the
following discussion.
7.2.1 Radial Forces
Radial forces are those that act in the radial direction and are
generated by the interaction of the current and the axial component
of the leakage flux density. They tend to squeeze the inner winding
and burst the outer winding. The calculation of radial forces does
not present much difficulty since the axial component of the leakage
flux density is calculable fairly accurately. Calculation of the stresses
due to this force acting on the conductors is more complex, especially
for inner windings. The compressive strength of winding is influ-
enced by the radial thickness of conductors, its work hardening
strength, number of blocks per circle, and winding dimensions etc.
Similarly, the outer winding experiences outward radial force
which results in tensile stress. The tensile strength of winding
depends on the work hardening strength of the conductors.
7.2.2 Axial Forces
Axial forces are those that act in the axial direction and are gener-
ated by the interaction of the current and the radial component of
the leakage flux density. These forces tend to bend the conductors in
the axial direction, and their sum total act on the coil-clamping ring
and other clamping structures. The calculation of axial forces
presents some problems since the radial component of the leakage
Figure 7.2 (a) Radial forces.

1. Inner winding 2. Outer winding.
(b)
(c)
Figure 7.2 (b) Axial forces.

(c) Most common pattern of radial component of
leakage flux.
flux density is difficult to calculate accurately. When ampere turns

are perfectly balanced so that the leakage flux pattern is symmetri-
cal, then the leakage field is axial over the major part of the coil
height. But due to the flux lines dispersing in the radial direction
in the vicinity of the winding ends, the axial flux density tends to
decrease, and the resultant flux density at the ends can be resolved
into the radial component causing axial forces. These axial forces
are unequally distributed between the outer and inner winding, due
to the presence of core.
The axial forces at the top and bottom are in opposite directions
as the currents are in the same direction. In case the ampere turns
are perfectly balanced and the leakage flux pattern is symmetrical,
the resultant force on the winding would be zero. Any axial displace-
ment between the magnetic centres of HV and LV windings will
result in a net axial force, tending to increase the displacement even
further. However, in practice, a complete balance of all elements of
the winding cannot be achieved entirely for a number of reasons like
provision to tappings, HV line lead exit from the centre of winding,
dimensional accuracy and stability of windings, etc.
7.3 Basic Formula and Methods for Force Evaluation
7.3.1 Basic Formula for Force Calculation
The basic equation for the calculation of any electromagnetic force

is
� � �
F = i ¥ B¥l (7.1)
where
�
B is the flux density due to leakage flux at mean radial
depth of winding,
i the short-circuit current, and
l the length of the conductor.
An examination of Eq. (7.1) shows that, since F is the cross-
product of i and B, a radial flux will give rise to axial force while an
axial flux will give rise to radial force.
The right hand side of Eq. (7.1) involves only three quantities,
viz. i, B and l. Of these, the current i and the length l are known
exactly. The flux density B is due to the leakage flux, since the
conductors lie in the field of the leakage flux. The calculation of the
radial and axial components of the leakage flux density is the most
difficult task and forms the crux of the problem.
7.3.2 Methods of Force Calculations
Attempts in solving the force problem were made by various

designers. The main thrust was towards obtaining a closed-form
solution, so that the designer would directly apply a set of formulae
and graphs to his particular design for calculating the forces. With
the advent of the digital computer, field plotting techniques were
developed and today numerical methods of force calculation using
the digital computer are common.
Calculation of radial forces requires evaluation of the axial
leakage-flux density. Formulae for the axial leakage-flux density
are readily available and are the same as used for leakage reactance
calculations. (Refer para 9.14 of Chapter 9). Hence, the radial forces
are easily calculated to good accuracy. On the contrary, evaluation
of the radial leakage-flux density, required for axial force calculation,
is quite complex. Various investigators have given different empirical
formulae and constants based on their experiments, for the calcula-
tion of the axial forces. Such formulae are to be used with due care,
as to their applicability to the case in hand.
7.3.3 Analytical Methods
(a) Billing1,2 has given formulae and curves for the calculation
of radial and axial short-circuit forces in transformers. His
formulae are derived considering the windings as current
sheets. These are fairly accurate for close-wound coils (spiral
coils).
(b) Waters3,4 has given formulae and constants for the calcula-
tion of radial and axial forces. He uses the residual ampere-
turn method to obtain the unbalanced ampere-turns from
which the axial forces are calculated using certain empirical
constants. The method is suitable for transformers with only
two distinct windings or coils.
7.3.4 Numerical Methods

The most well-known methods for force calculation in this category
are due to Roth5,6. Roth’s method is basically an analytical method.
He has used the method of images to obtain the magnetic flux
density as a double-Fourier series. In rectangular coordinates, the
solution is in the form of algebraic equations; in cylindrical coordi-
nates, the solution involves Bessel and Struve functions. The method
is most suitable for programming on a digital computer. A detailed
exposition of the method can be found in Ref. 7.
Rabins10 introduced a simpler form of the field solution using a
single Fourier series as opposed to the double-Fourier series of
Roth. Rabins developed this for reactance calculations, but the ex-
pression for the vector potential which he obtained, when differenti-
ated, leads to the flux density and hence the electromagnetic forces.
This method also is most suitable for digital computers, though the
basic formulation leads to an analytic expression. Other numerical
methods involve obtaining a point-to-point solution for the magnetic
field and then applying Eq. (7.1) systematically. Such field solutions
can be obtained in a number of ways; finite difference methods and
finite element methods are the most popular of these.
7.4 Radial Forces
A simple formula for the average radial force can be derived by

evaluating the axial component of the leakage flux density and
applying Eq. (7.1). An assumption is made that all the leakage flux
is axial only and passes between top and bottom yokes in straight
lines. This assumption is also made in calculation of leakage
reactance and will lead to a slightly higher value of radial force.
The flux density at the mean radial depth of the winding is given
by
1
B = m ¥ NI/L (7.2)
2 o
The average radial force is
Fr = B ¥ i ¥ l
1 NI
= m ¥ I ¥ pDm N
2 o L
1 ( NI ) 2
= mo ¥ ¥ pDm N
2 L
( NI ) 2
= 19.739 ¥ 10–7 ¥ Dm N (7.3)
L
Due to the assumption that all the flux is axial in nature,
Eq. (7.3) will result in a force larger than that due to any other
formula. It is used in all stress calculations and in the design of
windings. The maximum force per turn will occur at the inner turn
of innermost layer of the outer winding and the outer turn of the
inner winding and will be given by
2Fr
Fm = (7.4)
N
If the turn experiencing this force has small radial dimension and
is unsupported by any adjacent turns, the stress calculations are
done using Eq. (7.4). Otherwise the average force of Eq. (7.3) is used
for stress calculations.
7.5 Axial Forces
The electromagnetic forces arising due to a short circuit are oscilla-

tory in nature, and act on an electric system immersed in oil and
consisting of winding conductors, insulation components, and
clamping structure. Such forces, dynamically transmitted to various
parts such as conductors, end supports, press plate, and clamps may
be quite different, both in magnitude and in waveshape from
the internally generated electromagnetic forces depending on the

relationship between the excitation frequency and the resonant
frequency of the system.
When evaluating axial forces, winding misalignments, caused by
workshop tolerances need to be considered.
The design force calculations are performed both, for the sym-
metrical configuration and the displaced configuration of windings.
In case of a symmetrical winding arrangement in axial direction
having uniform current distribution, there is no resulting force
against the yoke, and the winding tends to be compressed in the
axial direction only.
Figure 7.3(a) is a typical case of a symmetrical winding arrange-
ment with no forces towards yoke. Different yoke distances and
tapping in the main winding and uneven current distribution in
the axial direction can cause the force integral to reach a final
value greater than zero. For example in Fig. 7.3(b), the residual
force, Fr is towards the bottom yoke, while in Fig. 7.3(c) it is seen
that the integral has passed the zero line once and obtained a
negative value at the bottom yoke resulting in residual force
towards top yoke.
Fmax Fmax Fmax
Fr Fr
(a) (b) (c)
Figure 7.3 (a) No forces on yoke.

(b) Force towards bottom yoke.
(c) Force towards top yoke.
While a simple formula could be obtained for radial forces, such is

not the case for axial forces. The two major questions encountered
are:
(a) What is the value of ampere-turns causing the radial
leakage flux?
(b) What is the effective length of path for the radial flux?
An approximation to the ampere-turns causing radial leakage-
flux is the residual ampere-turns caused due to tappings, gaps
in windings, unequal winding heights, etc. (see Fig. 7.4). This is
only an approximation, since even perfectly balanced windings will
still have fringing of flux at the coil ends, giving rise to radial
component of flux. The approximation is on the conservative
side, since the peak value of the residual ampere-turns are used for
calculations. Such a peak will not occur in practice, or will occur
only at one point.
(a) (b)
(c)
Figure 7.4 Unbalanced windings causing axial forces.

Considering the second question about the effective length

of path for the radial flux, even an approximation is not easily
forthcoming. The flux lines are influenced by the core legs, yokes,
the clamping structure, the tank walls and all other magnetic
structures within the tank, and follow quite complex paths for
closure. An exact evaluation of the effective length of path is
extremely difficult, if not impossible.
Attempts have been made to obtain empirical equations for the
effective length-of-path of the radial leakage-flux. Billig2 proposed a
value based on Rogowski’s work on sandwich windings. It is argued
that this gives wrong results in certain cases. Waters3 has given
values for the factor pDm/l for various cases. Forgestad8 has given
some empirical formulae for the effective length, based on his
experiments. Knaack9 has given certain other formulae from his
experiments.
The simplest of all these approaches is that of Waters3. Consider-
ing that there is an unbalance of NI ampere-turns, the mean radial
flux Br is given by
m NI
Br = o ¥ a ¥
2 leff
The axial force on the winding is
Fa = Br ¥ i ¥ l
mo NI
= ¥a¥ ¥ I ¥ pDmN
2 leff
mo pDm
= a (NI)2 N (7.5)
2 leff
Reference (3) gives the value of the factor (pDm/leff) for various
residual ampere-turn configurations and two values of the ratio of
window height to core circle. The paper opines that these two values
of the ratio suffice for all normal power transformers.
7.6 Roth’s Method of Force Calculation
A brief exposition of Roth’s method of force calculation is given here,

since it is an important and accurate method of evaluation. Roth’s
solution in rectangular coordinates is treated.
The transformer window region is considered and all the four
boundaries are assumed infinitely permeable. This rectangular region
contains an arrangement of a number of rectangular conductors (p)
carrying current, as shown in Fig. 7.5. The field solution is obtained

by using the method of images. Equation (7.1) is then applied to
obtain the forces.
cj
d
cj ¢
dj ¢
dj
x
c
Figure 7.5 Transformer window region with conductors.
From a consideration of the method of images, the p conductors

will form an infinite number of images both in the x and y directions.
Hence the field function must be periodic with both coordinates, and
will be given by
A = S S B1 cos mx . cos ny + S S B2 cos mx . sin ny

m n m n
+ S S B3 sin mx . cos ny + S S B4 sin mx . sin ny (7.6)

m n m n
Applying the boundary conditions and simplifying the vector
potential function will have the form
• •
px py
A= Â Â Bh, k cos (h – 1)
c
◊ cos (k – 1)
d
(7.7)
h=1 k=1
Equation (7.7) must satisfy Poisson’s equation if it is to be the

true field solution. Applying this condition the following relations
for the constant Bh,k are obtained:
(a) When h π 1
and k π 1
p
4m o 1 sin mh c ¢j - sin mh c j
Bh,k =
cd mh2 +
Â
nk2 j = 1
Jj
mh
sin nk d ¢j - sin nk d j
(7.8)
nk
(b) When h = 1 and k > 1
p
2m o sin nk d ¢j - sin nk d j
B1,k =
cd
Â Jj (c¢j – cj) nk
(7.9)
j=1
(c) When h > 1 and k = 1
p
2m o sin mh c ¢j - sin mh c j
Bh, 1 =
cd
Â Jj (d j¢ – dj) mh
(7.10)
j=1
(d) When h = k = 1, B1, 1 is a constant, which can be ignored
because A has an arbitrary origin.
In the above
p
mh = 2 (h – 1)
2c
p
nk = 2 (k – 1)
2d
To calculate the forces, it is necessary to evaluate the flux density
acting within the conductor. The flux density is simply the gradient
of the vector potential function of Eq. (7.7). This is evaluated and
the force calculated for a filament of area dx ¥ dy. This is then
integrated over the area of the conductor. The final equations for
the force is given by
• • (cos mh c ¢j - cos mh c j ) (sin nk d ¢j - sin nk d j )
Fxj = – Jj Â Â Bh,k nk
h=1 k=1
(7.11)
• •
(sin mh c ¢j - sin mh c j ) (cos nk d ¢j - cos nk d j )
Fyj = Jj Â Â Bh,k mh
h=1 k=1
(7.12)
Where Bh,k is given by Eqs (7.8), (7.9) or (7.10) as the case may be.
A fairly accurate solution is obtained when h and k are varied from
1 to 20.
7.7 Modes of Failure of Windings and Design of Windings
7.7.1 Failure Due to Radial Forces
(a) Outer Windings

The outer of two concentric windings will experience a simple hoop-
stress under short-circuit. This is so, since the outer winding has no
extraneous radial support and all the force is to be with-stood by the
conductors. The average force per turn from Eq. (7.3) will be
Fr
Frt = (7.13)
N
The pressure due to this force is
F 1
Prt = r (7.14)
N pDm ta
The hoop-stress due to this pressure is
Fr 1 Dm
Pht =
N pDm ta 2 tr
Fr
= N/m2 (7.15)
2 Nac
The above is true when there is more than one turn in the radial
direction and each such turn supports the other, so that all conduc-
tors may be assumed to have the same mean stress. This is not so if
the outer conductors do not support those with a smaller radius.
In such a case, the innermost conductor, nearest to the duct
will experience twice the average force as given by Eq. (7.4). It is
unlikely that an inner turn is subjected to bursting force and
remains unsupported by the turn over it. First, the coil is wound
under tension to form an integral and compact winding.
Secondly, a consideration of plastic behaviour will lead to the
conclusion that very slight plastic deformation will bring the second
and subsequent turns into the analysis. Thus, the mean stress is a
good indicator of the hoop-stress on the outer winding.
A recent paper by Patel12 shows that the maximum hoop-stress
may be about 20% higher than that calculated from Eq. (7.3). He has
applied the theory of elasticity to the problem, and by using a force
generation rate, has set up a second-order differential equation. The

solution yields the hoop-stress.
The designer has to ensure that the conductor section chosen is
adequate to withstand this hoop stress. If inadequate, the cross-sec-
tion has to be increased or a more work-hardened copper has to be
used.
(b) Inner Windings

A much more difficult problem is presented by inner windings,
whether disc or helical. Normally, inner windings are supported from
the core by spacers at a number of points along the circumference.
The simplest approach is to treat the conductor as a simply-supported
beam between two such spacers. This will give an erroneously large
value for the stresses. The winding resembles a tube supported
internally and subjected to external pressure. Failure can be of two
types: bending of the conductors between supports as in Fig. 7.6(a)
or by buckling as in Fig. 7.6(b). The designer must ensure that
neither of this occurs.
Eve
nly
forc distribu
the e along ted
win
ding
(a) (b)
Figure 7.6 Modes of failure of inner windings.
In general, the problem of the collapse of the inner winding is

difficult to solve. Solutions have been obtained considering the
winding as a tube supported internally and subjected to external
pressure, but others have experimentally proved these formulae to
be conservative. Treatment of the winding as a curved beam or in
general as a ring subjected to external pressure and supported
internally at a number of points has also been done. Kojima et al11

have presented a finite-element analysis of the problem and claim
good agreement with experimental results.
7.7.2 Failure due to Axial Forces

Assuming that the axial spacers are rigid and do not fail, axial
forces will tend to bend the conductors in an axial direction between
the axial spacers. Again, an approximate method is to treat each
segment as a simply-supported beam. Analysis as a continuous
beam can also be done.
A typical failure in disc coils is when
the conductors tilt in a zig-zag pattern
as in Fig. 7.7. Waters4 has given an
analysis of this type of failure, and his
expression for what he calls the critical
load Wcr is
pEc Nac ta
Wcr = (7.16)
6 R
The failure of windings due to axial
forces depends more on the strength of
the clamping structure and the insulat-
ing spacers. Once the clamping structure
gives way even slightly, the resultant
displacement of the winding enormously
increases the forces and lead to a com- Figure 7.7 Failure of disc
plete collapse of the winding. All manu- winding.
facturers of transformers use some kind of prestressing of the wind-
ings in order to obviate movement of coils due to short-circuits and
shrinkage of insulation spacers in service. The design of clamping
structures to with stand short-circuit force is a specialized area by
itself and is discussed in the following section.
7.8 Strengthening of Coils to Withstand Short-circuit Forces
In spite of adequate design strength of the conductor, a coil will still

fail under short-circuit, if it is permitted to move under the effect of
short-circuit forces. The very nature of these forces is such that they
tend to increase the unbalance causing them. Thus, any movement
of the coil greatly increases the force causing the movement and
ultimately leads to failure.
Some of the means adopted for strengthening the transformer

against short-circuit forces are given below:
Adequate radial supports are provided for inner coils.
Precompressed insulation material is used in manufacture in
order to minimize shrinkage in service.
Proper processing and prestressing of coils is done to obtain
dimensional stability of windings.
Depending upon prevalent design practices, devices like springs
or hydraulic dampers are used to absorb the impact load due to
short-circuit forces.
High tensile steel is also used in places where ordinary steel is
not adequate.
7.9 Design of Clamping Structures
The purpose of a well designed clamping structure (Fig. 7.8) is to

prevent any movement within the windings. This can be achieved
1 1
3 3
H.V. WNDG
H.V. WNDG
2
L.V. WNDG
L.V. WNDG
3 3
1 1
(a) Side view (b) Winding details

1. End frame
2. Clamp plate
3. Clamping ring
Figure 7.8 Typical construction.

by any clamping device which puts the coils permanently under a

pressure, higher than those produced by the short-circuit forces.
Thus, the first requirement for clamping structure is that, these
are designed for exerting short-circuit forces without permanent
deformation.
7.9.1 Stresses in Top Clamping Ring

For maintaining a constant pressure on windings, the force is applied
by a set of bolts or spring loaded jacks. These bolts are tightened
against top-end frame for providing the required pressure to windings
through top-end ring. Refer to Fig. 7.9 for top clamping ring. A, B, C
and D are the locations of bolts/jacks. Under short-circuit force, the
winding will try to bend the free span of the clamping ring
between the two consecutive bolts. The stresses in top ring are
A + B
C + D
Figure 7.9 Top clamp ring.
Bending moment
W12
M = (7.17)
10
Section modulus of ring
1 2
Z = bt (7.18)
6
Bending stresses
M
fb = (7.19)
Z
The results obtained by Eq. (7.19) are on the conservative side,
because of the assumption made for boundary condition of fixing
bolts as average of simply supported and fixed beam conditions. The
allowable bending stresses further depend on allowable factor of
safety, which is generally very low for short-circuit forces.
7.9.2 Stresses in Clamp Plates

During the period when short-circuit forces are developed, axial
forces are also developed in windings, which in turn transfer to top
and bottom end frame and these try to separate out from the core.
Hence, to keep them under position, clamp plates are used. These
also take the static load of transformer during lifting. The tensile
stresses are developed at section AA and shearing stresses at BB as
shown in Fig. 7.10.
(a) Tensile stress
P
ft = (7.20)
( X - D1 ) ¥ t
(b) Shearing stresses
P
fs = (7.21)
2 ¥ H1 ¥ t
(c) Bearing stresses
P
f¢b = (7.22)
D1 ¥ t
The stress values computed from Eqs. (7.20) to (7.22) should be
less than the yield point of the clamp plate material.
7.9.3 Stresses in Top-End Frame

The stresses in top-end frame are developed due to the following:
(a) The force due to the clamping pressure applied for tightening
the coils.
(b) The end thrust created by electromagnetic forces in the
windings, due to an unbalance in ampere turns.
(c) The lifting load of transformer.
B B
H1
A B B A
D1
X
H
B B
A A
D1
B B
Figure 7.10 Clamp plate.
In general, the short-circuit force condition is most critical.

Figure 7.11 shows the top-end frame. The end frame shown is for
single-phase transformer, where A is the core bolt position while C
is the pin used for connecting top-end frame with clamp plate. Since
the end frame is symmetrical along centre line, only half portion
shall be analyzed for stresses. Figure 7.12 shows the short-circuit
forces, acting locations and support position. The core bolt points
are assumed as support locations.
The bending moment at support is given by
Ms = F1 ¥ L1 + F2 ¥ L2 + F3 ¥ L3 + F4 ¥ L4 (7.23)
Core bolt
A A C Pin
+ + +
B D E
L1
L3
L2
L1
L
Figure 7.11 Top end-frame.
F3
F1
F4 F2
R R
Figure 7.12 Free body diagram.
The section modulus of end-frame is calculated by (see Fig 7.13)

Fig. 7.13 Section of end-frame
b1 ¥ h12 (b1 - t1 ) (h1 - 2 ¥ t1 )
Zs = - (7.24)
6 6 h1
The bending stresses are given as
Ms
fb = (7.25)
Zs
The maximum bending moment will be at C, where the force sign
is changing. Bending moment at C is given by
M c = F1 ¥ L1 + F2 ¥ L2 (7.26)
Section modulus is calculated from Eq. (7.24) and maximum
bending stresses should be less than yield point of end-frame mate-
rial with sufficient factor of safety.
b1
t1
h1
Figure 7.13 Section of end-frame
7.9.4 Bottom-End Frame

The same philosophy is applicable to the bottom-end frame. How-
ever, the flanges of bottom over which windings rest are critical
areas. Sufficient stiffening of flanges shall be done by providing
vertical stiffeners below flanges.
7.10 List of Symbols
All units are SI

m0 Permeability of free space = 4p . 10–7.
a Unbalance between windings expressed in per-
unit.
ac Cross-sectional area of turn.

A Vector potential.
B Flux-density at mean radial depth of winding.
Br Mean value of radial leakage flux-density.
B1, B2, UV Constants
B3, B4 W
Dm Mean diameter of main intercoil duct.
Ec Modulus of elasticity of conductor material.
�
F Force vector.
Fm Radial force on conductor adjoining the main
inter-coil duct.
Fr Radial force at mean radial depth of winding.
Frt Average radial force per turn.
�
i Current vector.
I Asymmetric peak of the short-circuit current.
l Length of conductor.
L Length of path for axial leakage flux.
leff Effective length of path for radial leakage flux.
N Number of turns in winding.
Pht Hoop-stress on turn.
Prt Average radial pressure per turn.
R Radius of turn.
ta Axial dimension of turn.
tr Radial dimension of turn.
Wcr Critical load.
Fa Axial force on the winding.
W Force per unit length due to short-circuit forces.
1 Pitch between two consecutive bolts in cm.
b Width of top ring in cm.
t Thickness of top ring in cm.
P Total short-circuit force on one clamp plate in
kg.
X Width of clamp plate in cm.
D1 Hole diameter for pin in cm.
H1 Height of clamp plate from centre of hole to top/
bottom surfaces in cm.
F1, F2, F3 and F4 Short-circuit forces intensities acting on end-
frame in kg.
L1, L2, L3 and L4 Locations of forces from centre line in cm.
REFERENCES
1. Billig, E., Mechanical Stresses in Transformer Windings, Critical
Resume, E.R.A. Report Q/T 101, 1943.
2. Billig, E. Mechanical Stresses in Transformer Windings, Journal
I.E.E., Vol. 93, Part II, June, 1946.
3. Waters, M., The Measurement and Calculation of Axial
Electromagnetic Forces in Concentric Transformer Windings,
Journal IEE, Vol. Part II.
4. Waters, M., The Short-circuit Strength of Power Transformers,
Book, Macdonald and Co, London, 1966.
5. Roth, E., Analytical Study of the Leakage Field in Transformers and
of the Mechanical Forces Acting On the Windings, General Electric
Review, Vol. 23, p. 773, May 1928.
6. Roth, E., Magnetic Leakage Inductance in Transformers With
Cylindrical Windings and Forces Acting on the Winding, RGE, Vol.
40, August 1936.
7. Binns, K.J. and P.J., Lawrenson, Analysis and Computation of
Electrical and Magnetic Field Problems, 2nd edition, Pergammon
Press Ltd., Oxford, 1973.
8. Fergestad, R., Electromagnetic Forces in Core-type Transformers
with Concentric Windings, Paper 114, CIGRE, Vol. II, 1956.
9. Knaack, W., The Mechanical Stressing of Transformer Windings
Under Short-Circuit, Paper 135, CIGRE, Vol. II, 1956.
10. Rabins, L., Transformer Reactance Calculations With Digital
Computers, Communications and Electronics, Vol. 75, p. 261, 1956.
11. Kojima, H., H. Miyata, S. Shida and K. Okuyama, Buckling
Strength Analysis of Large Power Transformer Windings Subjected
to Electromagnetic Force Under Short-Circuit, IEEE Trans. on PAS,
Vol. 99, No. 3, p. 1288, May-June 1980.
12. Patel, M.R., Hoop-Stress in Transformer Coils Under Radial Short-
Circuit Forces, IE(I), Electricial Engineering Division. Vol. 61, June
1981.
C HAPTER 8
Cooling Arrangements
C.M. Sharma
In power transformer, the oil serves a dual purpose as an insulating

medium as well as a cooling medium. The heat generated in the
transformer is removed by the transformer oil surrounding the
source and is transmitted either to atmospheric air or water. This
transfer of heat is essential to control the temperature within
permissible limits for the class of insulation, thereby ensuring
longer life due to less thermal degradation.
8.1 Various Types of Cooling
8.1.1 ONAN Type Cooling
The generated heat can be dissipated in many ways. In case of

smaller ratings of transformers, its tank may be able to dissipate
the heat directly to the atmospheric air, whilst bigger ratings may
require additional dissipating surface in the form of tubes/radiators
connected to tank or in the form of radiator bank. In these cases,
the heat dissipation is from transformer oil to atmospheric air by
natural means. This form of cooling is known as ONAN (oil natural,
air natural) type of cooling.
8.1.2 ONAF Type Cooling
For further augmenting the rate of dissipation of heat, other means

such as fans blowing air on to the cooling surfaces are employed.
The forced air takes away the heat at a faster rate, thereby giving
better cooling rate than natural air. This type of cooling is called
ONAF (oil natural, air forced) type of cooling. In this cooling
arrangement, additional rating under ONAN condition viz. after

shutting off fans, is available, which is of the order of 70–75%.
8.1.3 OFAF Type Cooling
Still better rate of heat dissipation could be obtained if in addition

to forced air, means to force circulate the oil are also employed. The
oil can be forced within the closed loop of transformer tank and the
cooling equipment by means of oil pumps. This type of cooling is
called OFAF (oil forced, air forced) type of cooling. Mixed cooling
transformers of radiator type can have two or three ratings avail-
able, one for each type of cooling, viz. OFAF, ONAF and ONAN.
OFAF cooling can also be obtained by using OFAF compact
coolers. These coolers offer the advantage that they occupy less space.
The disadvantage of OFAF compact coolers is that, there is no ONAN
rating available in case of failure of fans and pumps, thereby neces-
sitating provision of a stand-by cooling equipment for switching in,
immediately upon receipt of failure signal. Continuity of auxiliary
supply to fans and pumps is to be ensured for uninterrupted power
flow.
8.1.4 OFWF Cooling
Since the ambient temperature of water is always less than the

atmospheric air, it is possible to use water as a better heat-transfer
media. Such an arrangement employs oil to water heat exchangers.
A prerequisite for such an arrangement is the availability of a
source of sufficient quantity of water. In most of the transformers
for hydropower stations, this type of cooling is used. Such a cooling
is called OFWF (oil forced, water forced) type of cooling.
8.1.5 Forced Directed Oil Cooling
Additional means of improving the heat dissipation rate are

also employed on higher ratings for transformer. These comprise
arrangements which direct the transformer oil in the windings
through predetermined paths. This directed oil flow type of cooling
is utilized with advantage in case of forced oil system. The cool oil
entering the transformer tank from the cooler/radiators is passed
Cooling Arrangements 179
through the windings in a pre-decided manner ensuring faster rate

of heat transfer. This type of cooling is called ODAF (oil directed, air
forced) or ODWF (oil directed, water forced) type. Figure 8.1 shows
a typical arrangement used for directing the oil in the winding.
1. Winding conductors
2. Directed oil flow washers
3. Insulation cylinders
Figure 8.1 Directed oil flow.
8.2 Cooling Arrangements
Depending upon the type of cooling and rating of the transformer,

the cooling equipments can be arranged in various ways. Some of
the commonly used ones are described hereunder.
8.2.1 Arrangement with Radiators
Radiators are commonly used for ONAN, ONAF/ONAN and OFAF/

ONAF/ONAN types of coolings. Figure 8.2 shows a typical radiator.
Radiators consist of elements joined to top and bottom headers.
Radiator
flutes or
oil paths
1
3
1. Radiator headers
2. Radiator elements
3. Radiator disconnecting valves
Figure 8.2 Typical five flute radiator.
Elements shown in Fig. 8.2 are made by welding two previously

rolled and pressed thin steel sheets to form a number of channels or
flutes through which oil flows. These radiators can be either
mounted directly on the transformer tank or in the form of a bank
and connected to the tank through the pipes. The surface area avail-
able for dissipation of heat is multiplied manifolds by using various
elements in parallel. As oil passes downwards, either due to natural
circulation or force of a pump in the cooling circuit, heat is carried
away by the surrounding atmospheric air.
(a) Tank Mounted Radiators

Tank mounted radiators are used for smaller ratings of transformer
for ONAN or ONAF/ONAN type of coolings. The radiators are
connected to the tank with an interposing valve at the top and the
bottom. The hot oil enters the radiators from the top and after
dissipating the heat to surrounding air it goes back in the tank from
the bottom. The radiators are arranged on the tank in such a man-
ner that the centre line of radiator is always at a higher level than
the transformer winding centre line. This difference in the heights
is called thermal head and is responsible for producing thermo-
syphonic action, enabling the oil to circulate. Figure 8.3 depicts
the available thermal head in case of tank mounted radiators. It is
1 2
3
Centre line radiator

Available thermal head
Centre line coil
1. Transformer tank
2. Transformer core and winding
3. Cooling radiators
Figure 8.3 Available thermal head.
obvious that it is governed by the height of the tank and the length
of the radiator.
(b) Banked Radiators

Radiators can also be arranged to form a bank. This is necessitated
when transformer rating is higher and the total number of radia-
tors, required to dissipate the transformer losses, are more. Under
such circumstances it is not possible to physically arrange the radia-
tor on to the transformer tank. The radiators are mounted on to the
headers, which are supported from the ground. Interconnecting
pipework is required to be provided with isolating valves for main-
tenance and erection purposes. Separately mounted radiator bank
is generally called for in case of OFAF type of cooling, since the
interconnecting pipework is required to mount a pump for force
circulating the oil. Figure 8.4 (Plate 3) shows a transformer with
separately mounted radiator bank.
In case of tank mounted radiators, the available thermal head
cannot be adjusted much, whereas in case of banked radiators,
supports for the headers can be adjusted in height to give adequate

thermal head enabling better cooling under natural circulation also.
The cooling fans required for ONAF and OFAF types of coolings can
be arranged either underneath the radiators or on the sides.
The total number of radiators required for cooling of transformer
can be arranged in many ways. Usually one bank of radiators for
100% capacity is sufficient. However, if desired 2–50% banks can be
formed by dividing the number of radiators required into two equal
parts, each part being capable of dissipating 50% losses. Each 50%
part can be arranged in a completely independent bank with sepa-
rate auxiliary cooling equipments like fans and a pump. An alterna-
tive arrangement having 2–50% capacity radiators with one 100%
capacity pump is also specified. It is obvious, that because of the
additional pump and bank supports, etc., the fully independent al-
ternative shall be costlier.
To cater to the exigencies of failure of auxiliary pumps and fans,
provision of a standby pump and a fan is also sometimes specified.
In case of failure of the running pump and fan, the standby pump
and fan can be brought into operation. It is desirable to specify bare
minimum standby capacity, that too only for cases where continu-
ous operation at full load is envisaged. In a few cases, 3–50% radia-
tor banks have also been specified, but in our opinion one complete
50% bank as standby in case of mixed cooling banks is not neces-
sary. This is due to the fact that all mixed cooling transformers have
a self-cooled (ONAN) rating of at least 50% inherently available.
8.2.2 Arrangements Using Coolers
Compact OFAF coolers for dissipation of heat to atmospheric air

and OFWF coolers for dissipation to water are also in common use.
These coolers are connected to the transformer tank through inter-
connecting pipework and pumps for oil circulation.
(a) Oil to Water Heat Exchangers (OFWF Coolers)

This type of cooling equipment is commonly employed where water
is available in abundance. Figure 8.5 shows a cutaway section of
such a heat exchanger. Cylindrical shell (1) is generally made of
cast iron or mild steel which houses a removable tube stack (2). The
tube stack is generally made of copper, admiralty brass or stainless
5
1
4
1. Cylindrical shell 4. Water box
2. Removable tube stack 5. Safety expansion ring
3. Removable straight tube 6. Radial flow circular baffles
Figure 8.5 OFWF cooler cutaway section.

183
steel depending upon the impurities present in cooling water. The

tubes are assembled by roller-expanding the tube-ends in the
drilled holes of the tube plates. One tube plate is fixed between the
shell and water box (4). The other plate is free to move, to allows expan-
sion of the tube stack as a whole. A safety expansion ring (5) pre-
vents intermixing of oil and water. Water enters through water box
and flows through the tube stack while the oil flows over the tubes
guided by circular baffles (6) to give maximum cooling efficiency.
Water cooled oil coolers are available suitable for horizontal as
well as vertical mounting. In certain cases it is possible to mount
these coolers on the tank supported on brackets. Generally 2–100%
OFWF coolers are provided which includes one complete 100%
cooler as standby. It is very important to ensure that under working
condition, the oil pressure in the coolers in always more than the
water pressure, so that possibility of water leaking into oil is elimi-
nated. OFWF coolers can be arranged independent of each other.
But for better flexibility, usually both the coolers are specified to be
interconnected in such a manner, that by regulating the various
interposing valves it is possible to use either cooler with either of
the pumps. Such an arrangement is schematically shown in Fig. 8.6.
This arrangement requires two coolers and two pumps only.
(b) Oil-to-air Heat Exchanger (OFAF Coolers)

In this type of coolers, tubes with fins are employed to increase the
surface area. Figure 8.7 shows such a heat exchanger. The principal
components are ferrous or non-ferrous tubes threaded through
spaced aluminium or copper fins, to which they are metal bonded for
strength and heat transfer. These tubes are fitted into brass plates,
which in turn are attached to brass headers.
Axial-flow motor driven fans are attached to the cooler body rig-
idly to reduce vibrations. These coolers can be mounted either hori-
zontally or vertically. These coolers are mounted directly on trans-
former tanks in vertical position on small rating transformers.
These coolers are compact and as such occupy less space, and
therefore these are most suited where available space is constraint.
These coolers can be arranged in many combinations, viz. 1 ¥
100%, 2 ¥ 50%, 4 ¥ 25%, 5 ¥ 20%, etc. Usually, in addition to coolers
required for 100% rating, one complete unit cooler is kept as a
standby. A number of small unit coolers offer an additional advan-
Water inlet
Water outlet
W2 Pd W1
T5 T1
D2 T8 T4 T2 D1
T6
C2 C1
V8 V7
Pr4 T7 Pr3 T3 Pr1
Pr2
O2 V4 V3 O1
Common bus pipe
R2 R1
P2 P1
V6 V2 V1 V5
Tank
V — Valves W — Water flow indicator

R — Reflux valve D — Differential pressure gauge
P — Pumps Pr — Pressure gauge
C — Coolers T — Thermometer
O — Oil flow indicator Pd — Pressure reducing valve
Figure 8.6 2–100% OFWF cooller schematic.
tage, that under light load conditions some of the unit coolers can be
cut off, thereby saving the energy consumed by auxiliary fans and
pumps.
8.3 Propeller Type Fans
Fans constitute an important part of the radiator type cooling

equipment. The fans are suitable for outdoor duty and are required
to give large air deliveries at slow or moderate speeds with low
power consumption and noise level.
3
1
4
3
1. Cooler elements 3. Header

2. Cooling fan 4. Flange for connection
Figure 8.7 Cooler.
Figure 8.8 shows the main components of a fan with motor. The
fan blades are bolted on to the hub and the assembly called impeller
is properly balanced to minimise the wear of bearings, vibration and
noise level. The direction of air flow, also referred to as form of
running, is generally away from the motor. The blades are covered
1. Fan motor 3. Wire mesh guard

2. Fan blades 4. Fan support
Figure 8.8 Fan with its support.
from both top and bottom sides with wire guards, and the whole
assembly is supported on a steel structure when fans are mounted
below the radiators.
8.3.1 Fan Mounting Arrangements
Fans can be mounted underneath the radiators, blowing air from

the bottom to the top. This arrangement permits use of bigger sweep
fans. The air flow cone encompasses sufficient surface of the radia-
tors to give the required amount of heat dissipation. The radiator
surface not within the cone is natural cooled. These fans can be
supported either directly onto the radiators or mounted separately
from the ground. The latter case requires provision of foundations
for fan supports. A typical fan and support for ground mounted fan
is shown in Fig. 8.8 and fans mounted directly on to radiators are
shown in Fig. 8.9 (Plate 3). It should be ensured that fans mounted
on the radiators do not produce undue vibrations.
Another method of mounting the fans is on the side of the radia-
tors. This is with a view to ensure that the air blown cools elements
of radiators where oil temperature is higher. This arrangement is
used when a number of small capacity fans are used. To accommo-
date a number of fans, rows of fans are formed in the height of the
radiators. Figure 8.10 (Plate 3) shows such an arrangement.
8.4 Transformer Oil Pump
OFAF and OFWF type of cooling methods require oil circulating

pumps in the cooling circuit of a transformer. These pumps circulate
comparatively cold oil through the windings and carry away the
heat generated quickly, to keep hot-spot temperature in the wind-
ings within permissible limits. This is a closed-circuit operation and
the pumps are required to develop enough pressure, to overcome
the frictional head loss during the flow of oil in pipe work, cooling
equipment and winding, etc.
8.4.1 Constructional Features
Transformer oil pumps employ special features of construction, in

which the pump and motor is an integral unit. The pump impeller is
screwed and locked on the driving motor shaft. The motor unit is
embodied inside the pump casing, completely immersed in trans-
former oil, and is cooled by the surrounding oil. As the motor is
attached to the pump without the intervention of a stuffing box, the
equipment forms a leakproof glandless close-coupled unit suitable
for outdoor installation. No shaft seal is required, thus eliminating
the dangers of oil leakage or entry of air. The heavy-duty ball and
roller bearings carrying the shaft are lubricated by internal circula-
tion of transformer oil. The casing of pump is made of cast iron or
aluminium, with flanged suction and delivery for bolting to the
transformer oil-pipe circuit. The impeller is made of aluminium or
bronze.
Normally two different designs are employed depending upon the

duty requirements.
(a) Axial Flow-in-line Type

Single-stage axial-flow pumps are used to circulate large volumes of
oil against comparatively low frictional head losses. These pumps
are compatible with ONAN cooling mode of a mixed cooling (OFAF/
ONAF/ONAN) transformer employing radiator bank, as they offer
minimum thermosyphonic resistance to oil flow. Figure 8.11 shows
a cross-sectional view of such a pump.
3 4 7 5 8 6 1
2
1. Pump casing 5. Shaft
2. Window 6. Terminal board
3. Impeller 7. Stator assembly
4. Bearings 8. Rotor assembly
Figure 8.11 Cross-sectional view of an axial flow-in-line type.

(b) Radial Flow-in-line or Elbow Type

These pumps are used on transformers where oil is required to cir-
culate against large frictional head as encountered in oil-to-water
and oil-to-air heat exchangers. Figure 8.12 shows a cross-sectional
view of a radial flow elbow type pump.
Delivery
6 5
Suction
1 3 4
1. Pump casing 4. Shaft

2. Impeller 5. Stator assembly
3. Bearings 6. Rotor assembly
Figure 8.12 Cross-sectional view of a radial elbow type pump.
8.5 Flow Indicators
It is essential to monitor continuous flow of oil and water in the

cooling circuit of transformers for its safe working. Flow indicators
employed for this purpose perform the following two functions:
— indicate the full flow in proper direction.
— provide signal by operation of a switch when flow drops
below a preset limit.
8.5.1 Construction and Principle of Operation
A suspended vane is used as a sensor. Its surface is kept at right

angles to the direction of flow in the pipe. Figure 8.13 shows a sche-
matic diagram of a flow indicator. When liquid starts flowing
through the pipe, the vane gets deflected along with vane shaft.
This deflection is transmitted through a magnetic coupling to an
indicating needle on the dial. The needle indicates the flow of liquid
corresponding to vane deflection. When the flow in the circuit drops
to 70–80% of full flow, a switch provided in the instrument closes
with the help of a cam and initiates the alarm.
1 2 3 4 5 6
10 9 8 7
1. Pointer and dial 6. Vane

2. Mercury switch 7. Main pipe
3. Switch operating cam 8. Driving magnet
4. Return spring 9. Nonmagnetic wall
5. Vane shaft 10. Follower magnet
Figure 8.13 Schematic-diagram of flow indicator.
8.6 Stress and Hydraulic Analysis of Pipework
The analysis of piping may be divided into three categories:

(a) Stress analysis of pipe under internal pressure.
(b) Frictional head-loss analysis of complete piping system
including pipe bends, etc., under ONAN cooling.
(c) Frictional head-loss analysis of complete piping system,

including pipe bends, etc., under OFAF cooling.
8.6.1 Stress Analysis of Pipe
To check the leakage, etc., pipes are tested under very high oil pres-
sure. However, under normal conditions the internal pressure is
very low. Hence, to use the economical size and thickness of pipe, it
is imperative to do stress analysis of the pipe. The stresses in the
pipe under internal pressure may be computed from the following
formula.
PD
f = (8.1)
2t
It has been observed that the medium thin pipes are sufficient for
withstanding internal pressure.
8.6.2 Frictional Head Analysis of Pipes Under OFAF Cooling

A typical schematic diagram showing the complete piping system
with the fittings and pump is shown in Fig. 8.14. Under OFAF cooling
condition, the pump is required to work against a particular pres-
sure head and flow of oil for dissipation of heat through radiators. In
general, this pressure head may be divided into the following:
(a) Velocity head This is the height through which a body must
fall in a vacuum to acquire the same velocity as that of oil
flowing into the pipe and equals V2/2G.
(b) Entry head Head required to overcome the frictional
resistance at the entrance to the pipe.
(c) Frictional head This is due to the frictional resistance to
flow within the pipe. This includes pipe bends, etc.
(d) Frictional head in fittings This is due to the frictional
resistance to flow through the fittings, such as valves,
expansion joints, radiators, headers, etc.
(e) Frictional head in windings In the case of directed flow, the
oil is forced to circulate through the ducts in windings, etc.,
which in turn provides frictional resistance to flow.
All the above five elements of pressure heads added together give
the total head against which pump is required to generate the pres-
sure head. It is evident from pump characteristic, if the pressure
3
1 2
7
Elevation
5 6 1
1. Expansion joint
6 2. Filteration valve
3. Header
5 4. Radiator
5. Butterfly valve
6. Pump
7. Transformer
Plan
Figure 8.14 Schematic diagram of complete piping system.
head developed by pump is lower than the frictional head, the flow
of oil will reduce, which in turn will reduce the heat dissipation.
8.6.3 Frictional Head Analysis of Pipes Under ONAN Coolings

In the ONAN cooling system, the pump to force oil is not used. The
available thermal head due to top and bottom temperature and
dimensional difference is responsible for flow of oil under natural

cooling condition. This thermal head is also required to overcome
the frictional losses in the cooling system. The frictional head losses
are dependent on diameter of pipe, flow velocity and hence to decide
the correct size of pipe diameter, the overall study of head losses is
required to be done as illustrated for OFAF cooling.
8.6.4 Design Calculations (OFAF Cooling)

For finding out the total head-loss in the complete pipework, one
should know the type of flow which may be found out by calculating
the Reynold’s number. In 200 mm and 150 mm diameter pipes, the
Reynold number is very high and is within the turbulence zone.
Hence, the flow in cooling pipes is always turbulent. After finding
out the type of flow, the hydraulic losses in the following parts are
computed.
(a) Straight length of pipe
(b) 90° bends
(c) Valves
(d) Expansion joints
(e) Main header
(f) Sub-header
(g) Nipple
(h) Elements of radiators
(i) Due to contraction and enlargement of sub-header, nipple,
elements, etc.
(j) Pipe required for directing the oil into windings for directed
flow
(k) Exit from transformer tank
(l) Reducers, etc.
The frictional head-loss in straight length pipe would be
4 FLV 2
Hp = (8.2)
2 GD
The frictional head-loss in other components would be
V2
Hc = K (8.3)
2G
The factor K depends on many factors, such as diameter, Reynold
number and shape of components.
Some typical results are compiled and given in Tables 8.1 and 8.2
for OFAF and ONAN cooling respectively. Depending upon the
length of the pipework, number of bends and type of joints and other
fittings, the size of the pipework is selected for each type of cooling.
Depending on the velocity of oil in the pipe, the head losses in vari-
ous components may be computed from Eqs. (8.2) and (8.3), which
are summarized in Table 8.2. However, the driving force required to
make the oil flow inside pipe against frictional head is computed
from the following formula
d LT - d HT
Driving force = H ¥ (8.4)
d MT
The above given procedure is applicable to all types of cooling.
As most of the jobs have mixed cooling arrangements, in order to
decide the diameter of pipe and the optimum number of elements,
the ONAN cooling conditions become the governing criteria.
NOMENCLATURE
f Stress in kg/cm2
D Internal diameter of pipe in cm
P Working pressure in kg/cm2
t Thickness of pipe in cm
Hp Frictional head-loss in pipe length
Hc Frictional head-loss in components
F Frictional constant
L Length of pipe in metres
V Velocity of oil in pipe in m/s
G Gravity constant in m/s2
H Thermal head in cm
dLT Density of oil at lower temperature
dHT Density of oil at higher temperature
dMT Density of oil at mean temperature
Table 8.1 Summary of Pressure Head-Losses in Pipe Assembly in OFAF Cooling
196
Sl. No 20 Cms V Turbulent

Description Dia. Q Cm/Sec. or Remarks
GPM Laminar
1. Pressure head-loss in 1500 cms length pipe 20.0 600 145 Turb. 20 cm 15 cm
2. Head-loss in 90° bends (total 8 bends) 21.7 600 145 Turb. Reynold No.
3. Head-loss in butterfly valve (2 Nos.) 1.6 600 145 Turb. 36125 48172
4. Head-loss in expansion joints 9.9 600 145 Turb.
5. Main header 20 radiators with 28/36 elements 0.66 600 50 Turb.
6. Sub-header 28/36 elements 2.07 36 34 Turb.
7. Head-loss in nipples 0.86 1 10 Laminar
8. Head-loss in elements 0.26 1 2.4 Laminar
9. Head-loss due to contraction enlargement of
Transformers, 2/e
sub-head or nipple in radiators 0.11 1 100 Laminar

10. Head-loss at joint of sub-header to mark header 0.75 1 34 Turb.
11. Enlargement from pipe to tank — —
12. Contraction from 20 cm to 15 cm — —
Total 67.91 cm 810.4
For 20 cm dia.
RS Without directed flow 67.91 cm
T With directed flow 67.91 + 1.53 + 28.1 = 97.54 cm
Table 8.2 Summary of Pressure Head-Losses in Pipe Assembly in ONAN Coolings
Sl. No. Description 20 cm Q Remark

dia. GPM
1. Pressure head-loss in 1500 cm length pipe 0.821 121.9 With directed flow
2. Head-loss in 90° bend (total 8 bends) 1.3767 121.9
3. Head-loss in butterfly value (2 Nos.) 0.06 121.9
4. Head-loss in expansion joints 0.40 121.9 For 20 cm
Head-loss = 41.23 cm
5. Main header 0.0244 121.9
6. Sub-header 0.058 6.07
7. Head-loss in nipple 0.0397 0.217
8. Head-loss in elements 0.009 0.217
9. Head-loss due to contraction/enlargement of sub-heads,
nipple in radiator 0.006 0.217 Extra loss due to
direct flow = 3.36 cm

10. Head-loss at joint of sub-header of mainheader 0.0336 0.217
11. Head-loss in entering the tank 0.4 121.9
12. Reducer from 20 cm to 15 cm — 121.9
3.1507 cms
For 20 cm dia.
RS Without directed flow 3.1507 cms
T With direct flow 3.36 cms
197
C HAPTER 9
Design Procedure
R. C. Agarwal
9.1 Specifications of a Transformer
Transformers have large number of variables and their require-

ments differ a great deal. Complete specifications as stipulated in
Chapter 15 must be furnished for designing the transformer to suit
the specific requirements. In the following paragraphs design
procedure, selection of various parameters and their effect on the
various performance criteria have been discussed.
9.2 Selection of Core Diameter
Core diameter of a transformer depends upon a number of factors

like rating, percentage impedance between windings, basic insula-
tion level, transport height, overfluxing requirements, type of core
and quality of core steel. Hence, it is fairly complicated to derive a
universal and exact formula for determining diameter. In practice,
the core diameter is selected by the designer keeping into view the
limiting conditions as mentioned above. Based on this, guaranteed
parameters, viz percentage impedance and losses are worked out.
Core diameter is adjusted to meet the guaranteed parameters.
9.2.1 Influence of Varying Core Diameter
Increasing core diameter increases area of cross-section, thereby

increasing voltage per turn, which reduce the number of turns in
various windings. The percentage reactance between windings is
directly proportional to number of turns and diameters of various
Design Procedure 199
coils and is also inversely proportional to volts per turn and coil
depth. In order to have specified reactance, increased core diameter
necessitates reduction in coil depth and increase in coil dimensions
in lateral direction, which leads to reduction in core height and
increase in core leg centres. Inspite of reduction in core height due
to increase in core diameter, overall weight of core steel increases,
which also increases no-load loss of transformer. Also, reduced
number of turns in windings even with larger length of mean turn
results in reduced copper weight, which in turn also reduces the
load loss of transformer. Similarly, reduced core diameter results in
reduced core steel weight and no-load loss and increased copper
weight and load loss.
9.2.2 Core Area
Stepped core construction is used to obtain an optimum core area

within circumscribing circle of a core. Core area depends upon
number of steps, grade of core steel, insulation on the laminations,
i.e. varnished or unvarnished, and method of core clamping. As the
number of steps increases, core area also increases, however it
needs extra labour to cut the various sizes of laminations. Optimum
number of steps are used to give overall economy. Large rating
transformers are usually provided with high tensile steel clamp
plates for clamping the core laminations, which provide increased
core area for a particular core diameter. Depending upon the flux
density in the core, adequate number of ducts are provided to keep
its hot-spot temperature within permissible limits. (For more
details on hot-spot temperature and cooling ducts in core, refer to
Chapter 4).
9.3 Selection of Flux Density
Value of flux density is chosen to suit the guaranteed performance.

Normally flux density is chosen near knee point of magnetization
curve, however, adequate margin should be kept to take care for
system conditions like overfluxing, frequency and voltage varia-
tions. In certain cases value of flux density is reduced to limit the
noise level of transformers.
9.3.1 Influence of Varying Flux Density
Keeping the other parameters same, increase in the value of flux

density in the core results in higher volts per turn. Hence number of
turns in various windings are reduced. Effect of increase of flux
density on reactance is similar to that of increase in core diameter.
In order to meet the requirement of specified reactance, coil depth is
reduced and lateral dimensions of coils are increased. Inspite of
small increase in core leg centres, reduced core height results in
lower core-steel weight. Increased flux density, with corresponding
reduced core-steel weight results in higher no-load loss of trans-
former. Also reduced number of turns in windings results in lower
copper weight and load losses. Similarly, reduction in the value of
flux density causes increased core-steel weight, lower no-load loss
and increased copper weight and load loss.
9.4 Selection of Type of Core
The cores are made of CRGO steel, which is available in various

grades having different properties. Suitable grade of steel is chosen
to suit the performance requirement and overall economy. Different
types of cores are used as described in Chapter 4. Selection of type
of core depends upon the rating of transformer and transport limita-
tions. For large three-phase transformers, five-limb construction is
adopted to overcome the problem associated with the transport
height. For single-phase transformers, one centre-wound limb with
two return limbs is a common configuration, though sometimes
both limb wound cores are also adopted. For very large ratings, core
construction having two wound limbs with two return limbs may
also be adopted.
9.5 Selection of Leg Length
Maximum leg length available for designing the windings, is depen-

dent upon maximum transport height, type of wagon to be used for
transportation and type of core selected. By proper shaping of the
transformer tank for well wagon, the leg length can be increased.
Generally for larger rating transformers maximum value of leg
length gives the overall economy. Whereas, lower rating trans-

former may have reduced leg length to give economical design. By
reducing leg length, reactance of transformer increases, and vice
versa.
Influence of core diameter, flux density and leg length over
various parameters while maintaining the % reactance as constant
is tabulated below:
Parameter Increased Increased Increased

core flux leg
diameter density length
Copper weight Decrease Decrease Increase
Load loss Decrease Decrease Increase
Core steel weight Increase Decrease Decrease
No-load loss Increase Increase Decrease
9.6 Selection of Type of Windings
The windings along with its insulations form the electric circuit of
the transformer. Due care must be taken while designing the wind-
ings to ensure its healthiness during normal as well as fault condi-
tions. The windings must be electrically and mechanically strong to
withstand both over-voltage under incidence of surges and me-
chanical stresses during short-circuit conditions. The temperature
of windings at rated, over-load and short-circuit conditions should
be within the limits, ensuring the proper life of transformer.
The power transformers are manufactured for a very wide range
of outputs and voltages and realization of these requirements is pos-
sible only by using different kinds of windings. The following types
of windings are used in power transformers.
(a) Spiral winding
(b) Helical winding
(c) Reversed section winding
(d) Parallel layer winding
(e) Tapered layer winding
(f) Interleaved disc winding
Spiral winding is a medium-current and low-voltage winding.
Tertiary winding, of star/star/delta connected power transformer
used for stabilizing purposes and sometimes feeding small loads,

could be generally spiral winding.
Helical winding is a high-current and low-voltage winding.
Normally, it is used for LV coils of large generator transformers.
When number of turns preclude the use of helical winding, the
reversed section winding is used. It is generally used for high-
voltage and low-to medium current rating. Reversed section (disc)
winding is usually used up to 132 kV class windings.
Higher voltage windings above 132 kV class are mostly multi-
layer or interleaved disc winding. Number of layers of layer-type
windings are generally five to nine. Ordinary reversed section disc
winding is not suitable for voltage above 132 kV because of impulse
distribution characteristics of winding. However, by doing the inter-
leaving of conductors/turns, impulse distribution characteristics
improves and therefore, interleaved disc windings are used for HV
coil above 132 kV class.
Constructional details of various types of windings have been
dealt within Chapter 5.
9.7 Selection of Tapchanger
To cater for the voltage regulation in the transformer and system

voltage variation, off-circuit/on-load tapchanger is provided. Off-
circuit tap switch is provided when tapchanging is required only
occasionally. Generally generator transformers are provided with
off-circuit tap switch. When tapchanging is required under loaded
conditions, on-load tapchangers are provided. Generally, unit auxil-
iary transformers, station transformers, and system transformers
are provided with on-load tapchangers. The choice of tapchangers is
governed by the following factors:
(a) Tapping range
(b) Number of steps
(c) Step voltage
(d) Current rating
(e) Location of tapping
(f)Design of tapping winding, i.e. linear, reversing, coarse and
fine
(g) Insulation level
(h) Type of voltage variation, i.e. constant flux, variable flux and
mixed type
(i) Power flow requirement, i.e. unidirectional or bi-directional
Constructional details and various types of tapchangers have
been described in Chapter 6.
9.8 Calculation of Number of Turns
emf equation of transformer can be written as under:

Volts per turn,
Et = 4.44 Bm Af (9.1)
where Bm = Maximum flux density in tesla
A = net cross-sectional area of core in m2
f = frequency in Hz
Therefore, after fixing the core area and flux density, volts per
turn is calculated by Eq. (9.1). Then LV (lowest voltage winding)
turns are calculated. LV turns are rounded off to nearest integer
and volts per turn is adjusted and new value of flux density is
worked out by back calculation. The number of turns in other wind-
ings are determined by the new value of volts per turn.
9.9 Selection of Conductor and Current Density
Phase values of currents for different windings are calculated.

Selection of current density in any winding depends upon type of
winding, loss level and temperature rises permitted. Current
density in any winding should be such that winding gradient and
hot-spot temperature are within limits. Sometimes, current density
in a winding is restricted due to short-circuit-force withstand point
of view. Generally current density in various windings is chosen
so as to meet the load-loss requirement and also such that their
gradients are close to each other, thus resulting in an economical
cooling equipment.
After deciding the values of current densities, copper area
required for each winding is determined. Now suitable number of
parallel conductors or pretransposed cables in both radial and axial
direction and their width and thickness are selected, such as to
meet the desired performance values. For large current ratings,

pretransposed cables offer the advantage. Conductors of chosen
dimensions must be strong enough to withstand short-circuit forces;
also they should not have excessive eddy currents.
9.9.1 Influence of Varying Current Density
Increased current density reduces copper weight but leads to higher

load loss, gradient and forces in the winding. Reduced copper weight
results in lower core dimensions causing reduction in core-steel
weight and no-load losses.
Transformer windings have insulation within the winding, between

windings and windings to earth. Insulation within the winding is
generally paper insulation, however helical and disc type of winding
have duct between turns or discs. Paper thickness of conductor
should be such that it should be able to withstand various voltage
stresses appearing during normal and transient conditions. Some-
times paper thickness is increased on pre-transposed cables of large
cross-sectional area, to increase the mechanical strength of paper
insulation. Electrical clearances between windings of various voltage
class and windings to earth depend upon their BIL and insulation
arrangement adopted. Various clearances and disposition of solid
insulation should be such that adequate cooling ducts are available
to have effective cooling of windings. Also voltage stresses are
controlled within limits.
Details of insulation design, method of improvement of voltage
stress, etc., have been discussed at length in Chapter 5.
9.11 Calculation of Lateral and Axial Dimensions of Coils
Based upon number of parallel conductors in radial direction and

covered thickness of conductor, lateral dimensions of various coils
can be calculated. If there are a number of layers in a coil, the
lateral dimension of layer can be worked out and by taking into
account the gap between layers, total lateral dimensions of coil can
be worked out. Axial dimension of a coil depends upon covered axial

dimensions of conductors and total number of conductors per turn in
axial dimension. If cooling ducts are provided between conductors,
these are also to be accounted for. Extra space is provided for
dummies and transpositions if required.
9.12 Ampere-turn Balancing
HV and LV winding are disposed axially in such a way that perfect

ampere-turn balancing throughout the length of coil is achieved.
Location and type of tapping coil disturb the ampere-turn balancing
at different tap positions. Inter-wound type of tapping coil, in which
each step occupies the whole axial length of winding, is ideal from
ampere-turn balance point of view. Ampere-turn balancing between
windings is essential from short-circuit force point of view.
Unbalanced ampere-turn may give excessively high forces. Also, if
ampere-turn balancing is not proper, it may give considerable
increase in stray losses due to radial component of leakage flux.
9.13 Reactance Calculation
The estimation of reactance is primarily the estimation of the distri-

bution of the leakage flux and the resulting flux linkages with
primary or secondary coils. The distribution of the leakage flux
depends on the geometrical configuration of the coils and of the
neighbouring iron masses and also on the permeability of the latter.
The diagrams in Fig. 9.1 show typical leakage flux distribution. In
case of cylindrical-core type coils of equal length [Fig. 9.1 (a)], the
leakage field is packed in the space between HV and LV windings
and it runs parallel to the core for nearly the full length of
coils. Where there is an inequality in the coil lengths, the field is
considerably altered, as shown in Fig. 9.1(b).
9.13.1 Cylindrical Concentric Coils (Equal Length)
For this case the actual leakage field, e.g. Fig. 9.1(a) is assumed to
consist of a longitudinal flux of uniform and constant value in the
(a) C.L. of window
(b) C.L. of window
Figure 9.1 Leakage flux.
inter-space between primary and secondary, and a field crossing

the conductors, reducing linearly to zero at the outer and inner
surfaces. Further, the permeance of the leakage path external to the
coil length is assumed to be so large as to require the expenditure of
a negligible mmf, i.e. all the mmf is expended on the coil length. The
effect of the magnetizing current in unbalancing the primary and
secondary ampere-turn equality is neglected. Considering the Fig.
9.2(a), percentage reactance (X) between HV and LV winding can be
calculated by the following equation:
X =
FG
2p fm 0 Lmt ( AT ) b + b2
a+ 1
IJ ¥ 100 (9.2)
Lc Et H 3 K
where
f = Frequency (Hz)
m0 = Magnetic space constant = 4p ¥ 10–7
Lmt = Mean length of turn of primary and secondary
together (metre)
AT = Ampere-turns per limb of either coil
Lc = Coil length (metre)
Et = Volts per turn
a = Gap between HV and LV windings (metre)
b1 = Lateral dimension of HV winding (metre)
b2 = Lateral dimension of LV winding (metre)
b1 a b2 b¢1 a¢ b¢¢1 a b2
H.V. L.V. Lc H.V. H.V. L.V.
AT AT
1
AT
2
(a) (b)
Figure 9.2 Ampere-turn diagram for cylindrical coils.
When HV winding is split as in Fig. 9.2(b), percentage reactance

can be calculated by the equation
X =
FG
2p fm 0 Lmt ( AT ) b¢ + b2 + b1¢¢ a¢
a+ 1 +
IJ
¥ 100 (9.3)
Lc Et H 3 4 K
where b¢1 and b≤1 stand for lateral dimension of layers of HV wind-
ings and a¢ is gap between two layers.
9.13.2 Cylindrical Concentric Coils (Unequal Length)

The leakage field depends on the proportional difference in length,
and on where the difference occurs (e.g. at one or both ends or in the
middle, etc.). The case is very common, and it may be produced by
end-turn reinforcement, tappings or by the normal small difference
of coil length. The effect of divergences on the reactance requires to
be investigated for each individual case. Figure 9.3 shows one
KAT KAT
KAT
AT AT AT (1-K) AT KAT
(a) (b) (c)
AT
Figure 9.3 Analysis of asymmetrical cylindrical coils.
method of treating the problem. The coils, each of ampere-turns

(AT) are shown in the section. The actual arrangement (a) can be
considered as equal to the sum of (b), a symmetrical system of longi-
tudinal ampere-turns amenable to the treatment leading to Eq.
(9.2), and (c) an asymmetrical transverse system of ampere-turns
producing a cross flux. Having determined the flux distribution, due
to each of the systems (b) and (c), the reactance is determined from
the emf’s induced in actual windings due to each flux distribution.
9.13.3 Sandwich Coils
Figure 9.4 shows the simplified case where the end coils have
one-half the turns of the remainder. There will be n HV coils, (n – 1)
LV coils of equal ampere-turns and two LV half-coils. Percentage
reactance between HV and LV can be calculated by the equation
X =
p f m 0 Lmt ( AT ) FG a + b
1 + b2 IJ ¥ 100 (9.4)
n
W Et H 6 K
b1 a b2
Æ
Yoke
Figure 9.4 Leakage of flux in sandwich coils.
The percentage reactance can be calculated, for any coil arrange-

ment, with the help of ampere-turn diagram, as discussed above.
Desired value of reactance can be achieved by suitably altering the
various parameters.
The percentage impedance and reactance are practically the
same for large transformers, as they have very low winding resis-
tance. The regulation in the system fixes the highest acceptable im-
pedance and the requirements of transformer protection, which
fixes the lowest acceptable impedance value. It is of interest to note
that variation in impedance affects the cost of transformer. It is
difficult to give a specific picture of how transformer economics are
affected by variation in impedance, because this depends upon the
relative price of copper and core material, as also the rates at which
core-loss and copper-loss are capitalized. In general, for a given
material content, a high impedance value would result in higher
copper-loss and lower iron-loss. In any case, very high impedances
are undesirable, as they may give substantially high stray losses,
resulting in hot spots. Low impedance values are desirable where
regulation is regarded as most important and on the other hand,

they pose considerable problems for designing the transformer
suitable for large mechanical forces developed under short-circuit
conditions. Also, low impedance transformers would require large
core diameter, resulting in higher iron-loss, lower copper-loss and
higher transport weight of the largest packet.
9.14 Iron Weight and Losses
Transformer no-load loss occurs because of flow of main flux in the

core. It depends upon grade of steel, frequency, flux density, type
and weight of core and manufacturing techniques (i.e. mitred or
interleaved joints, bolted or boltless, core construction, etc.). Modern
core-steel used has got different magnetic properties in the direction
of grain orientation and cross-grain direction. Cores are designed
such that flux travels mostly in the direction of grain orientation to
give maximum advantage. However, at joints, flux travels in cross-
grain direction, giving increase in losses. For accurate estimation of
no-load loss, core weight for cross-grain and along-grain portion are
calculated separately. Steel characteristics furnished by their sup-
pliers give the iron-loss in watt/kg for both direction of flux travel
(i.e. along the grain orientation and across the grain orientation)
No-load loss is computed with these parameters.
9.15 Copper Weight and Losses
Winding conductor weight depends upon the specific gravity of con-

ductor material, number of turns in the winding and cross-sectional
area of conductor and this can be calculated by the following
equation:
G = DALN (9.5)
where G = weight of conductor material (kg)
D = density (kg/m3) = 8900 kg/m3 for copper
A = area of cross-section (m2)
L = mean length of winding turn (M)
N = number of turns.
By putting the corresponding values for different windings in

Eq. (9.5), their conductor material weights can be worked out.
The alternating current flow, in the windings, is associated with
load-loss comprising I2R-loss and eddy current loss in the windings
and stray losses in tank, clamping gear, etc. The resistive loss in the
windings, i.e. I2R-loss depends on resistivity of conductor material,
number of turns, and cross-sectional area and length of mean turn
of conductor.
The resistance (R) of winding is calculated by the following
equation:
r LN
R = (9.6)
A
where r is specific resistivity of conductor material in W-metre.
By knowing the resistance of each winding, I2R-loss in various
windings can be worked out. Eddy current loss in conductors de-
pends upon their configuration and stray loss in tank and clamping
structure, etc., depends upon leakage flux, and can be suitably esti-
mated with reasonable accuracy. The load-loss of transformer is a
function of temperature and generally expressed at a reference tem-
perature of 75°C.
9.16 Stray Losses in Transformer
The load losses in the transformer consist of losses due to the ohmic
resistance of winding and stray losses. These stray losses take place
in the loaded and unloaded windings, clamping framework and
transformer tank and depend upon leakage flux and magnetic field
surrounding the leads. The total stray loss in the transformer is
generally made up of the following components.
(a) Stray loss in the transformer tank due to leakage flux from
the windings.
(b) Stray loss in the tank due to leads.
(c) Stray loss in the clamping framework due to leakage flux
from the windings.
(d) Stray loss in the clamping framework due to leads.
(e) Stray loss in the loaded windings.
(f) Stray loss in the unloaded windings.
9.16.1 Leakage Flux

Some amount of leakage flux will always exist in all types of
transformers. The total leakage flux increases approximately as the
square root of MVA rating of transformer. Therefore, increased
capacity of transformer results in larger values of leakage flux and
this gets further amplified with increased value of reactance of
large transformers, causing increased stray losses. This leakage
flux can sometimes lead to local overheating. By controlling the
leakage flux path, stray losses can be reduced considerably, and
thus resulting in reduced local heating.
9.16.2 Calculation of Leakage Flux

Leakage flux is generated by the winding while carrying current
and is the resultant of the field surrounding the windings. This
leakage flux is carried in the gap between HV and LV windings. The
leakage field in the gap is fairly constant along the length of wind-
ings but at ends of winding, fringing of flux takes place. Various
paths followed by leakage flux are shown in Fig. 9.5
The component of flux parallel to the axis of coil is termed as
axial flux and that perpendicular to the axis of coil is termed as
radial flux. The axial and radial flux densities at various regions
can be calculated by various methods such as image method, Roth
method, Rabins method, etc.
9.16.3 Losses in Metallic Parts due to Leakage Field

As shown in Fig. 9.5, leakage flux cuts the various metallic parts
namely tank walls, core clamping plates, etc., resulting in eddy
current losses in them. These eddy current losses and the losses in
the region of high intensity field can lead to local overheating. The
regions of these losses and overheating are, the tank walls near the
winding ends, core clamping plates at winding ends, clamp plates
on core limbs, etc.
Leakage flux going back to core, does not give significant
losses and temperature rise. However, the flux impinging upon
flat surface of core, where the flux cuts perpendicular to the lamina-
tion, additional losses will take place and can lead to high local
temperatures.
The losses due to leakage field emerging from winding depend
upon the amount and intensity of flux, permeability and resistivity
Tank
A
Core
LV HV C
A B
A. Core clamp plate

B. Base plate
C. Flux density greatest in this region
of tank surface
Figure 9.5 Principal leakage flux path.
of metallic materials. Low resistivity materials will give rise to

higher eddy currents and higher losses. The losses also depend upon
the distance of metallic parts from the source of leakage flux, as
with the increased distance the amount and the intensity of flux is
reduced.
9.16.4 Stray Losses in the Winding

Major portion of stray losses takes place in the winding. This con-
sists of eddy current loss and circulating current loss. Eddy current
loss is proportional to the square of the flux field in which winding
conductor is lying and to the power of four of the thickness of the
conductor perpendicular to the direction of flux, and also to the

square of total number of radial conductors in the winding.
In large transformers, a turn in winding consists of large number
of conductors. In such an arrangement, unequal voltages are induced
in different conductors due to varying leakage field in radial direc-
tion and variation in length of each conductor. The unequal voltages
in parallel strands (conductors) give rise to circulating currents. The
losses due to circulating current within a few stands of turn may not
reflect much in overall losses, but still can cause over-heating of
strands.
Eddy current and circulating current losses can also take place
in idle windings, not connected in load circuit but lying in leakage
field.
9.16.5 Stray Losses due to Current Carrying Leads

Leads carrying current, induce magnetic field around the leads. If
field is strong and in close vicinity to metallic parts, it produces
eddy currents and losses. Intense field can lead to local hot-spots.
The resultant losses are not much important, but the hot spots can
lead to gasification of oil or decomposition of insulation. The effect of
current in the leads depends upon the magnitude of the current,
distance of leads from metallic parts and resistivity of material.
9.17 Stray Loss Control
A minimum stray loss design can be achieved by analyzing system-

atically, the source of leakage flux and its path. Main leakage field
in the transformer will always exist and losses arising out of this
can be reduced by various methods given below:
9.17.1 Magnetic Yoke Shields

Magnetic shields, made up of core laminations are used under yokes
as shown in Fig. 9.6. A large proportion of the axial leakage flux is
fed back into the yokes. The yoke clamp assembly is shielded and
reduction of radial flux to tank side is also achieved. Magnetic
shunts can be conveniently used for three-phase, five limbed and
single-phase three, limbed transformers. For three-phase, three
limbed transformers, the flux collection from outer limbs and
feeding to yokes is difficult and needs special transfer technique.
Core yoke
Top shield
Bottom shield
Core yoke
Figure 9.6 Magnetic yoke shields.
The reduction in stray loss is considerable in the case of large

transformers.
9.17.2 Magnetic Shunts

The magnetic shunts consisting of packets of core laminations,
are fixed inside the tank to absorb stray flux. The thickness of
lamination packets is decided by the flux density used. The typical

arrangement of magnetic shunts is shown in Fig. 9.7. The reduction
in stray loss in tank is considerable in large transformers.
Magnetic Tank
shunt
Coil Coil Coil
Figure 9.7 Fixing of magnetic shunts on three-phase transformer.
9.17.3 Electromagnetic Screens

High conductivity materials like aluminium or copper are used as
screens on tank walls. The stray flux induces eddy currents in these
screens, which suppresses the original field and relieves the tank
walls from stray flux. The screens are effective beyond a critical
distance and placement of screens within critical distance may
result in increased losses.
9.17.4 Losses in Clamp Plates and Core Packets

The core limb clamp plates are very near the leakage field area and
are subjected to intense radial field at the ends of windings. This
results in eddy losses on clamp plate ends and can lead to local
hot-spots. Stray loss and heating is further increased due to high
current inner winding leads. The outside packet of lamination on
core limbs are also subjected to similar conditions. The control
in temperature rise and losses can be achieved by using clamp
plates of non-magnetic materials like stainless steel. Substantial
reduction in eddy losses and temperature rise can be achieved by

providing slots on clamp plate.
The clamp plates made of thin lamination stacks and placed
edgewise are ideal for large transformers. Outer packets of core are
also sometimes subdivided to reduce temperature rise on core plates
for large rating transformers.
9.17.5 Losses in Winding
Subdivision of conductors radially reduces the eddy current loss due

to axial leakage field. Similarly, subdivision of conductors axially
reduces the eddy current loss due to radial component of leakage
field. To reduce radial component of leakage flux, it is essential that
ampere-turn of HV and LV windings shall be perfectly balanced,
which in turn gives the reduced eddy current loss due to radial
component of leakage flux. It is worth mentioning here that large
unbalance in ampere-turns may lead to very high stray loss. To
eliminate the circulating current between parallel strands (conduc-
tors) of a turn, transposition is essential. This leads to positioning of
strands of a turn, such that flux linkage is same, thus equalizing the
induced emf in each strand. Methods of various transposition have
been discussed in Chapter 5.
The use of transposed cables for high-current windings result in
considerable reduction in stray losses. In high-voltage winding with
moderate current, requiring two or three conductors in parallel,
bunched conductors can be used to improve winding space factor,
and to provide subdivision of conductors.
9.17.6 Losses due to Leads

Losses due to high-current leads can be reduced by spacing them
suitably from metallic structures. The field effect of leads can be
eliminated by positioning together the current carrying leads of
opposite direction (go-return arrangements). In three-phase connec-
tion, the leads of all the three phases can be grouped together so
that the net vectorial effect of field is minimum. Losses due to leads
can be reduced by shielding the nearby surfaces by non-magnetic
material like aluminium. High-current bushing mounting plate has
high eddy current losses. This can be reduced by putting the
non-magnetic inserts to break eddy current path or by using the
plate of non-magnetic material like aluminium or stainless steel for
mounting the bushings.
9.18 Impulse Calculation
The voltage distribution in power frequency-voltage tests is

substantially uniform between turns and coils. Under impulse test,
however, the distribution can be far from uniform and initial
voltage distribution is governed solely by capacitance network. It
depends upon value of a, which can be expressed by the following
equation:
Cg
a = (9.7)
Cs
where
Cg = winding to ground capacitance
Cs = series capacitance of winding
The determination of initial voltage distribution is merely the calcu-
lation of the various capacitances of windings. The greater the value
of a the greater is the divergence from uniform voltage distribution
from line to ground. It is evident that value of a is high for a small
transformer. For transformers of higher rating and especially for
higher line voltage, Cg decreases because it is determined largely by
clearances, while Cs increases because of greater radial depth of the
winding. Thus, a is lower and initial voltage distribution is more
close to uniform. Due to stress concentration at the ends of winding,
a few end turns are provided with extra insulation. To reduce
stress concentration at the ends, disc coil windings are provided
with stress rings that act as radial shields, although they do not
materially improve the axial distribution. Axial improvement can
be gained by the addition of shields or by several other means of
controlling the electric field distribution. Also refer to Chapter 5,
Clause 5.2 for surge behaviour of windings.
9.19 Mechanical Forces in Windings
When transformer is loaded, the flow of currents in its primary and

secondary windings gives rise to leakage flux and mutual forces
between windings. The mechanical repulsive force with normal
load currents is low compared to the strength of coils. Under fault
conditions, the forces produced may be increased many times. The
design of transformer windings, insulating cylinders, coil clamping
rings, clamping plates, etc., should be such that they are able to
withstand these forces. The forces produced are radial as well as

axial. Radial component of force tends to burst outer winding
and crush inner windings and axial component gives bending and
compressive stresses. Various forces and their effect have been
discussed at length in Chapter 7.
9.20 Temperature Gradient and Cooling Calculation
Difference between mean winding temperature and mean oil tem-

perature is termed as winding gradient. It depends upon the losses
in the winding and surface area available for cooling. Winding de-
sign, i.e. selection of current density, size of cooling ducts, etc.,
should be proper, resulting in low gradients.
Gradient of the windings can be reduced by providing the
directed oil flow, in the case of forced oil cooled transformer. The
gradient in various windings of a transformer are adjusted by
suitably altering the winding design to result in economical cooling
equipment and thus overall economy. The design of cooling equip-
ment is to be such that it shall be capable of delivering the rated
MVA, under specified conditions, without exceeding the guaranteed
values of top oil temperature rise and mean winding temperature
rise. Top oil temperature rise depends upon mean oil temperature
rise and half of the difference in oil temperatures at the inlet and
the outlet of the cooling equipment. Mean winding temperature rise
depends upon mean oil temperature rise and winding gradient.
Various types of cooling have been discussed in Chapter 8.
9.21 Typical Design Calculations for

Two Winding and Auto-transformers
Example 9.1 Typical design calculations for a 20 MVA, 132/33

kV, star/star, three-phase, 50-Hz, power transformer with on-
load tapping on HV winding for voltage variation of LV from – 5%
to + 15% in 16 steps. Percentage impedance 10% on 20 MVA base.
Select, core diameter = 515 mm, say
Bm = 1.524 tesla, say
Type of core: Three limbed construction
A = 0.18167 m2 (Core utilization factor 87.2%)
f = 50 Hz.
Substituting the values in Eq. (9.1),

Et = 61.46
33000
LV turns per phase (N1) = = 310
3 ¥ 61.46
1,32,000
HV turns per phase (N2) = = 1240
3 ¥ 61.46
1240
HV turns (max.) = = 1305.3
0.95
= 1305 (say)
1240
HV turns (min.) = = 1078.3
1.15
= 1078 (say)
Tap winding turns = 1305 – 1078 = 227
(Linear tapping winding foreseen)
H.V. winding design

Type of winding: Reversed section winding
Number of coils in parallel
= 2 (mid-entry of line lead)
Number of discs in each coil
= 50
Conductor size = 2.1 ¥ 5.4 (mm)
Radial paper insulation over conductor
= 0.60 mm
Cooling duct between two discs
= 6.50 mm (compressed)
Maximum number of turns in a disc
= 22
20,000
Phase current =
3 ¥ 132
= 87.48 A
Current density = 3.98 A/mm2 (say)
Conductor area = 21.96 mm2
Axial dimension of coil

0.60 ¥ 2
=
1.20
10% paper compression – 0.12
1.08
5.40
6.48 ¥ 50
324.0
Cooling ducts 6.50 ¥ 49 = 318.5
642.5 ¥ 2
1285
Top and bottom insulation
235
Leg length 1520 mm
Radial dimension of coil

0.60 ¥ 2
=
1.20
2.10
3.30 ¥ 22
72.6
Say 74 mm
Tapping winding design

Type of winding: Reversed section winding
Number of coils in parallel = 4
Number of discs in each coil = 32
= 0.60 mm
Cooling duct between two discs
= 3.70 mm (compressed)
Maximum number of turns in a disc = 9
Copper area in coils = 22.88 mm2 (for current density of
3.82 A/mm2, (say)
Axial dimension of coil

Calculating in the same manner as for HV coil, axial dimen-
sion/coil for 32 sections = 252.00
The coils shall be placed concentrically one over the other.
Radial dimension of coil

For 9 turns per section, calculating in the same manner as
for HV, radial dimension of coil = 29 mm (inclusive of building
tolerance)
LV winding design
Type of winding: Reversed section winding (single coil con-
ceived)
Number of discs = 110
= 0.225 mm
Cooling duct between two discs = 3.70 mm (compressed)
Maximum number of turns in a disc = 3
20,000
Phase current = = 349.92 A
3 ¥ 33
Copper area in coil = 103.65 mm2
Current density in coil = 3.38 A/mm2
Calculating as for HV coil the axial dimension of coil
= 1520 mm (inclusive of 235
mm top and bottom insula-
tion)
Radial dimension of coil = 50 mm (inclusive of building
tolerance)
Reactance calculation at normal tap (Fig. 9.8)

K ( AT )p
%X = ¥ S DM ¥ 10–5 (9.8)
EtL
K = constant
= 3.86
L = reactance length
= 1368 (including fringing effect)
S DM LV = 10,117
Gap = 30,844
HV = 24,055
=
3.86 ¥ p ¥ 349.92 ¥ 310 ¥ 65016 ¥ 10 - 5
%X =
61.46 ¥ 1368
= 10.2 (Guaranteed impedance 10%)
Core weight and loss calculations

Weight of legs = 7.65 ¥ 3 ¥ 1816.7 ¥ 152 ¥ 10–3 = 6337
Weight of yokes = 7.65 ¥ 4 ¥ 1816.7 ¥ 107.5 ¥ 10–3 = 5976
Weight of corners = 7.65 ¥ 2 ¥ 1816.7 ¥ 50 ¥ 10–3 = 1390
= 13,703 kg
LV HV Taps
Core dia
515
21 50 44 74 44 29
1.0 1.0
0.1306
Figure 9.8 Ampere turns diagram at normal tap position.
At 1.524 tesla, specific watt/kg of M4 steel

= 0.9176
No-load loss = 1.15* ¥ 0.9176 ¥ 13,703 ¥ 10–3
= 14.5 kW
Copper weight calculations

LV = 8.89 ¥ p ¥ 607 ¥ 310 ¥ 103.65 ¥ 10–6 = 544.8
HV = 8.89 ¥ p ¥ 819 ¥ 1078 ¥ 21.96 ¥ 10–6 = 541.7
Taps = 8.89 ¥ p ¥ 1010 ¥ 227 ¥ 22.88 ¥ 10–6 = 146.5
1233.0 kg
* Building factor from experience of similar transformers.

Total weight for three phases = 1233.0 ¥ 3

= 3699.0 kg
Resistance calculations
LV = 21.05 ¥ p ¥ 607 ¥ 310 ¥ 10–6/103.65 = 0.1201 W
HV = 21.05 ¥ p ¥ 819 ¥ 1078 ¥ 10–6/21.96 = 2.6590 W
Taps = 21.05 ¥ p ¥ 1010 ¥ 227 ¥ 10–6/22.88 = 0.6627 W
Total HV resistance at normal tap
162
= 2.6590 + 0.6627 ¥
227
= 3.1319 W
Copper loss calculations at normal tap

LV = (349.92)2 ¥ 0.1201 ¥ 10–3 = 14.7
HV = (87.48)2 ¥ 3.1319 ¥ 10–3 = 24.0
38.7 kW
Total I2R loss for 3 phases = 116.1 kW
Eddy current loss and stray losses based on similar past design
as 5 kW and 10 kW respectively
Total copper loss = 131.1 kW
Impulse distribution across different sections of coils, tem-
perature gradients of windings and mechanical forces are further
done to ascertain the design adequacy vis-a-vis test require-
ments.
Example 9.2 Typical design calculations for 20 MVA, 132/66/11

kV, Star/Star/Delta, 3 phase, 50 Hz auto transformer with onload
tappings on HV winding for voltage variation of HV from – 10% to
+ 10% in 16 steps. Percentage impedance 12.5% on 20 MVA, base.
Core diameter – 405 mm, say
Bm = 1.684 Tesla, say
Type of core: Three limbed construction
A = 0.11145 sq. metre (core utilisation factor 86.5%)
f = 50 Hz
Substituting the values in Eq. (9.1),
Et = 41.66
11000
LV turns per phase (N1) = = 264
41.66
66000
IV turns per phase (N2) = = 914.7
13 ¥ 41.66
Say 914.5
132000
HV turns per phase (N3) = = 1829.4
3 ¥ 41.66
Say 1829
Common Winding turns = 914.5
Series Winding turns = 1829 – 914.5
= 914.5
Trapping turns = 1829 ¥ 0.1
= 182.9 Say 184 (reversing taps)
Current calculations
20,000
LV phase current =
3 ¥ 3 ¥ 11
= 202.02 A
20,000
IV phase current =
3 ¥ 66
= 174.96 A
20,000
HV phase current =
3 ¥ 132
= 87.48 A
Series winding phase current
= 87.48 A
Common winding current = 174.96 – 87.48
= 87.48 A
Series and common windings can be reversed section coils.
Tapping winding can be inter-wound spiral coil placed between
common and series windings or elsewhere depending upon the
impedance variation required. LV winding can be spiral coil.
Design of these coils can be done in a similar manner. While
calculating impedance and other performance, auto factor is to be
used suitably as applicable.
REFERENCES
1. Say, M.G., “Electrical Engineers Reference Book”.
2. “The J & P Transformer Book”.
3. Blume, L.F., A. Boyajian, G. Camilli, T.C. Lennox, S. Minneci,
and V.M. Montsinger, “Transformer Engineering” a Treatise on
the Theory, Operation and Application of Transformers”, Book,
Chapman and Hald Ltd., London, 1951.
4. Say, M.G., “Performance and Design of AC Machines”, Book, Sir
Isaac Pitman & Sons, 1962.
5. Narke, D.V., and R.K. Talwar “Paper on Stray Losses in Power
Transformer” BHEL, Bhopal.
6. Khosla, A.K., Paper on “Recent Trends in the Design of Large Size
Transformers” BHEL, Bhopal.
C H A P T E R 10
Structural Design
of Transformer Tank
M.K. Shakya
S. G. Bokade
Transformers are housed in metallic tanks, which are structurally

robust enough to withstand the loadings such as full vacuum during
processing of transformers, oil pressure and concentrated point
loads of lifting, hauling, jacking, etc. The tank sizes reach the trans-
portable limits and call for a lot of ingenuity in the design to meet
such stringent conditions as minimum electrical clearances from
high voltage points of windings and leads, and proper shaping to
reduce oil quantity, transportable profile suitable for loading on rail
wagons, transportable weight, etc. From these design consider-
ations, in general, and for large power ratings, transformer tanks
are structurally quite complicated. Though, for medium size, plain
tanks are also used quite often for the sake of ease, and economy of
cost of fabrication.
The structural design of transformer tanks comprises the compu-
tation of the combined behaviour of plate and shells with stiffeners,
which involves a realistic estimate of boundary conditions. For com-
puting the stresses and displacements at a few selected points the
classical method is handy, however for computing the stresses and
displacements in global sense, the classical method is not sufficient
and one has to make use of rigorous methods such as finite element
method.
10.1 Types of Tank Constructions
Rectangular tanks are simpler in fabrication. However for large

rating transformers, shaping of tanks becomes necessary to conform
to transportable profile. Shaping is provided by rounded corners at

the ends, truncation of lower portion of walls from consideration of
loading in well wagon girder and on the covers to reduce the height.
To minimize the tank oil, the tank profile may closely follow the
electrical clearances along the coils. As is evident, shaping gives
saving in tank material and oil but increases complexity and fabri-
cation cost. From the standpoint of structural strength, shaping is
advantageous and this aspect is discussed later in the text.
Transformer tanks may be classified as–
— plain tanks
— shaped tanks
— bell-shaped tanks
— corrugated tanks
— stub end type tanks
10.1.1 Plain Tanks (Fig. 10.1)
Plain tanks are rectangular box type in shape. These tanks may
sometimes also have rounded corners. These are commonly used for
small and medium rating transformers. This type of construction
may however be uneconomical for larger sized tanks, particularly if
the welding is done manually.
10.1.2 Shaped Tanks (Fig. 10.2)
In order to make more economical transformers, it is possible to

shape the profile of the tank body suitably, so that the inside volume
of the tank is less. It is suitably truncated in the lower portion for
ease of loading inside the well of the wagon and it also ensures the
required minimum electrical clearance at all points. Usually many
shaped tanks have the tank walls shaped towards HV-leads side,
where electrical clearances are comparatively much larger, and a
flat wall towards the LV side. Structurally the curved portions in
the tank walls act as stiffener, which is an added advantage.
The shaping in the tank construction is decided by the electrical
layout considerations of transformer windings and terminal gear/
tapchanger mounting arrangements.
Structural Design of Transformer Tank 229
10.1.3 Bell Shaped Tanks (Fig. 10.3)
Sometimes it is desirable to construct the tank in two parts, so that

if the top portion is removed, the height of the lower part is such
that the core and the winding of the tank are easily accessible for
inspection and maintenance. This arrangement is preferable at site,
where it is not possible to lift the core and windings together from
tank for inspection and maintenance work. Such tanks which are
made into two separable parts are known as bell type. The profile of
the tank walls can be either rectangular or bell shaped.
10.1.4 Corrugated Tanks (Fig. 10.4)
An alternative for providing vertical ribs welded to the plates is to

form corrugation on the plates by suitably folding the plates. These
corrugations play the role of stiffeners and this process reduces
welding by replacing the welding vertical ribs on the tank walls.
The usual type of tank construction with corrugated plates for tank
walls employs box type of horizontal stiffeners to provide necessary
structural strength against bending stresses. Another advantage of
corrugated tank walls is that, it provides additional cooling area on
the tank walls and tank weight reduction.
10.1.5 Stub-end Wagon Type Tanks (Fig. 10.5)
It may not be possible to transport, very large rating power trans-

formers, on any conventional wagon. The height of the girder on
which the consignment rests adds to the overall transport height.
As a remedy, such large-size transformers are not supported on
such girders. The tanks are suitably designed such that they can be
Transformer tank
Figure 10.5 Stub-end type tank.

supported from either end by stub-end wagons and the transformer

hangs in the vertical position, leaving the minimum necessary
clearance between the bottom of the tank and railway track. These
tanks are of special construction and designed to withstand the dy-
namic loads during transit in addition to static loads.
10.2 Structural Design of Transformer Tanks
10.2.1 Classical Method
Structural design is necessary to guarantee against material failure

under design loads. For the sake of clarity and ease of formulation,
the analysis has been provided for a plain rectangular tank. The
whole tank structure is suitably subdivided into a number of plate
panels. The shaped transformer tanks can be similarly subdivided
in such plate panels and the present analysis with some approxima-
tion would also be valid for shaped tanks.
For the plates with stiffeners in either direction (i.e. transverse
and longitudinal), it has been seen that the behaviour of trans-
former tank wall is very near to the average behaviour between sim-
ply supported and encasterated beams.
10.2.2 Theoretical Formulation
Simplified rectangular tank with horizontal box stiffeners and ver-

tical ribs (stiffeners) is illustrated in Fig. 10.6. For the purpose of
analysis the entire face is subdivided into a number of panels,
whose boundaries are edges of the face and centre lines of stiffeners.
This is illustrated in Fig. 10.7. These plate panels are subjected to
design loads like vacuum, oil pressure, concentrated point loads,
etc.
10.2.3 Role of Stiffener
Figure 10.8 shows the rectangular plate supported all around. For
such a plate loaded under uniformly distributed load (udl) p consid-
ering a simply supported beam along longitudinal direction (i.e., in
the direction of l), the bending moment at the centre is given by
Figure 10.6 Rectangular tank with horizontal and vertical stiffeners.
Panel 5
Panel 2 Panel 3 Panel 4
Panel 1
Figure 10.7 Subdivision of tank wall into panels.

t
I
l
h
Figure 10.8
pbl 2
M = (10.1)
8
The stresses and deflections may be computed by the following for-
mulae
1 2
Zp = bt (10.2)
6
M
fss = (10.3)
Zp
pbl 2
= (10.4)
8 ¥ Zp
5 pbl 4
d= (10.5)
384 EI
Referring to Fig. 10.9, the combined section modulus of plate and
stiffener will now be taken in the above formula as
Z = Zp + Zs (10.6)
I = Ip + Is (10.7)
Consequently, the maximum stresses and the deflection will
be proportionately reduced and these can be computed from the
above formulae by substituting the combined section modulus and
moment of inertia.
Box stiffener
Combined C.G.
Plate
Figure 10.9 Combined section plate with stiffener.
Similarly, the bending moment, stresses and deflections for a

encastered beam may be given as
pbl 2
M = (10.8)
24
pbl 2
fEB = (10.9)
24. Z
pbl 4
d= (10.10)
384. EI
By considering stiffeners as shown in Fig. 10.9, the combined
section modulus of plate and stiffeners will be taken as in Eq. (10.6).
Consequently, by substituting these values of section modulus
and moment of inertia, the values of stresses and deflection will
come down proportionately. In case of transformer tank, the walls
have some average behaviour of the two cases, namely simply
supported and encastered.
f = ( fss + KfEB)/2 (10.11)
d = (dss + KdEB)/2 (10.12)
where K is an empirical value and will depend on the weld size, the
rigidity of the corners and on stiffener sizes.
10.2.4 Behaviour of Continuous Plates with

Intermediate Flexible Supports
The bending of each span of the plate can be computed by combining
the known solution for laterally loaded, simply supported rectangu-
lar plates, with those of rectangular plates bent by moments
distributed along the edges.
The expressions for stresses and displacements of plates
and plates with stiffeners, considering bending effect in both the
direction, are given as
F I
pb l 2 G
G 1 JJ
f1 = 3
(10.13)
12 I G
GH 1 + K II FH bl IK JJK
xx
1
yy
xx
F I
plb G
G
2
1
JJ
f2 = (10.14)
GH 1 + K II FH bl IK JJK
12 I G yy
3
xx
1
yy
F I
f¢1 =
pb G
2
G 1 JJ (10.15)
2 4
2t G
GH 1 + K FH bl IK JJK
1
F I
pl G2
1 JJ
f¢2 =
2t G
G
2 4
(10.16)
GH 1 + K FH bl IK JJK
1
Subscript 1 for shorter side and 2 for longer side have been used
in the above expressions.
F I
pl b G
G 1 4 JJ
d¢ = (10.17)
K EI G
GH 1 + II FH bl IK
2 yy xx
4 JJ
yy K
F I
d≤ =
pl G
G
4
1 JJ (10.18)
3 4
K Et G
GH 1 + FH bl IK JJK
3
where Eqs. (10.17) and (10.18) are for the deflection of stiffened and
unstiffened plate respectively.
In case of actual transformer tanks, the dimensions are fairly
large and unless one uses very thick plates, one has to stiffen the
different faces by welding stiffeners of suitable sizes, either along
longitudinal or transverse or both the directions, such that the
resulting panel widths are safe values. The different arrangements

are illustrated in Figs. 10.10 – 10.13 and these are self-explanatory.
It may be pertinent to mention here that for pressure-loadings on
transformer tanks, the horizontal stiffeners are more effective as
these bear most of the bending moments compared to vertical stiff-
eners. But horizontal stiffeners alone may not be sufficient except
for smaller size tank, and a combination of these stiffeners along
with vertical ribs are most effective and most economic from design
standpoint. Here the variable parameters are the plate thickness,
sizes of horizontal and vertical stiffeners and the cost of fabrication.
In the case of curved tank structures, the vertically curved
surfaces act like vertical stiffeners. However, the bending moment
distribution along the profile of the tank walls are very complicated.
H H1
Figure 10.10 Structural considerations in tank design.
H2 H H1
Figure 10.11 Structural consideration in tank design.

H H1
H2 H H1
As such, shaped tanks are more complicated for structural analysis

by classical methods. However, some simplified assumptions can
be made based on experience, such as the resulting sub-division
of plate panels, the boundary conditions along the edges of these
panels and the formulae discussed earlier can be applied without
significant loss of computational accuracy.
10.2.5 Finite Element Method

Any analytical method is bases on certain simplifying assumptions,
like a particular type of boundary condition such as simply sup-
ported or fixed. The panel behaviour is further analyzed as a beam
problem or a plate problem. It is difficult to account for structural

irregularities, such as cutout or openings or bracket type of projec-
tions. Some of these zones may be areas of high stresses and need to
be investigated. It is very difficult to account for such complexities
in the tank structure by any classical analytical method. Neverthe-
less, it is necessary to investigate the stress contours in the global
sense, in order to improve reliability in the design and parametric
optimization based on the results of stress mapping under design
loads. A rigorous and more accurate method of stress and displace-
ment analysis is based on the formulation of finite element method,
which is discussed in the following section.
In finite element formulation, the region of interest is subdivided
into subregions known as finite element. Within each subregion
the variables are assumed in the form of a function. The functions
assumed for unknown variables are defined by interpolation func-
tions associated with the nodal values of the unknown parameters.
Using the interpolation functions, the variables at any point in the
subregions can be expressed in terms of their nodal values as
u = N de (10.19)
where u = vector of variables for which functions are assumed
N = matrix for interpolation functions
de = nodal unknown parameters for an element.
The stress vector s is related to the strain vector e by
s =De (10.20)
where the D matrix depends on the material properties like modu-
lus of elasticity, Poisson’s ratio, etc.
The strains are related to the nodal displacement by
e = B de (10.21)
This matrix B depends on the selection of the elements, viz. trian-
gular, quadratic, etc., curvature of the surfaces, the direction co-
sines of the local elements with reference to global coordination, etc.
The strain energy can be expressed as
f = ÚÚÚvol e T s dvol (10.22a)
T
= ÚÚÚvol B DBde dvol (10.22b)
Introducing the concept of stiffeness matrix K as
Fext = Kde (10.23)
From the principle of virtual work, Eq. (10.23) may be equated
with Eq. (10.22b), and we get
Ke = ÚÚÚBT D B dvol (10.24)
Kg1oba1 = ÚÚÚBTTDTTB dvol (10.24a)
p = Kd (10.25)
From Eq. (10.25), the nodal displacement can be computed. From

the nodal displacement the stresses can be computed as
s =De
= D B de (10.26)
For the overall analysis, the stiffness matrix (K) of all the elements
are assembled appropriately. For the assembly of the elements,
the basic data needed are the element—node relationship and nodal
displacement relationship.
10.2.6 Choice of Element

Several types of elements have been developed by the researchers
working in the area. The isoparametric series possesses distinct
advantages over others by virtue of ease of their applicability to
curved surfaces.
For analysis of tank, the plates/sheets are idealized by eight
noded quadrilateral isoparametric elements, and stiffeners by three
noded quadratic elements. These elements are found to be most
suitable.
One of the salient features of formulation of these elements is
introduction of thickness vector of the plates/shells and stiffeners
and to allow the rotation of the normals, which in turn accounts for
the bending and shear stresses in the plates and stiffeners. The
formulation of these elements provides the simplicity of analysis of
two-dimensional plates without any compromise on the accuracy of
results
Another significant aspect of this formulation is the introduction
of stiffener elements and its merging with the plate elements. This
sophistication has been incorporated without adding any complex-
ity in the formulation or computation.
10.2.7 Formulation of Eight Noded Isoparametric Shell Element

In finite element method formulation, the part which is to be
analyzed is subdivided into numbers of element variables, with the
region of element related to the nodal unknown variables through
interpolation function, as given by Eq. (10.19).
The interpolation function is known as shape-function. When the
same shape-function is used for defining the geometry as well as
displacements of the element, then the element is known as
isoparametric element. To account for the curved surfaces in the
tank structure, the curvature parameters x, h, z have been intro-
duced along there directions and these are connected to the local X,
Y, Z coordinate system of the elements by the following matrix

(Jacobian) relationship. (See Fig. 10.14).
Figure 10.14 Coordinate system.
R| ∂f U| R| ∂x ∂y ∂z U| R ∂f U
|| ∂x∂f || || ∂∂xx ∂x
∂y
∂x
∂z
|| ||∂∂xf ||
S| ∂h V| = S
||∂h ∂h ∂h V| S| ∂y V| (10.27)
|| ∂f || ∂x
|T∂z
∂y ∂z || || ∂∂fz ||
T ∂z W ∂z ∂z WT W
or {f, xhz} = [J ] { f, xyz} (10.27a)
The shape functions for eight noded quadrilateral isoparametric
elements are (Fig. 10.15)
1
Ni = (1 + xxi) (1 + hhi)/(xxi + hhi– 1) (10.28)
4
for i = 1, 3, 5
1
Ni = (1 – x 2) (1 + hhi) (10.29)
2
for i = 2, 6
1
Ni = (1 + xxi) (1 – hi2) (10.30)
2
for i = 4, 8
where x, h are the natural coordinates of point of interest and xi, hi
are the nodal coordinates of the node.
7
(–1, 1)
5
(1, 1)
6
(0, 1)
8
(–1, 0)
4
(1, 0)
1
(–1, –1) 3
2
(0, –1) (1, –1)
Figure 10.15 Two-dimensional eight-noded quadrilateral elements.
Once the shape functions are assumed, the next step is to arrive
at expressions for shape fields and displacement fields for element;
which are given by the following expressions in terms of thickness
vector of element
R|X U| 8 R|X ci U| 1 8 R|V xi U|
S|Y V| = Â Ni S|Yci V| + 2 z Â Ni S|V yi V| (10.31)
i =1 i=1
TZ W TZ ci W TV zi W
where X, Y, Z, are coordinates of a point in element Xci, Yci, Zci, coor-
dinates of a centre point in element
Vxi, Vyi, Vzi, thickness vector in X, Y, Z, direction
N, shape function
u, v, w, displacements
a, b, rotations
z, natural coordinates in third direction
mi, matrix of direction cosines in V1 and V2 vectors
R|u U| 8 R|u ci U| 8 R|a U| i
S| v V| = Â Ni=1
i S| v
ci V| + z /2 Â N t [m ] S V
i=1
i i i
|Tb |W
(10.32)
TwW Tw ci W i
10.2.8 Formation of Stiffness Matrix

The stiffness matrix is obtained by equating internal and external
work done for a vertical displacement, in terms in matrix form.
The internal work done

= ÚÚÚe s dvol (10.33)
On simplifying Eq. (10.33) the internal work done
= ÚÚÚ B de DB de dvol (10.34)
Similarly, external work done
= deTFext (10.35)
By equating expressions (10.34) and (10.35), and expression for
Fext will be
Fext = ÚÚÚBTDBde dvol (10.36)
Fext = kde (10.37)
Hence,
K = ÚÚÚvol BTDB dvol (10.38)
Expressing (D) in global coordinates and dvol in curvilinear coor-
dinates Eq. (10.38) is given as
K = ÚÚÚvol BTTT DTBJ/dx dh dz (10.39)
10.2.9 Nodal Forces

The nodal forces are given by
F = ÚÚÚvol NTP dvol (10.40)
where P = external force per unit volume
10.2.10 Formulation of Three Noded Isoparametric

Stiffener Element
To match with the quadratic shell element, an appropriate qua-
dratic stiffener element (Fig. 10.16) shall be used. It can be easily
combined with the shell element after accounting for its eccentricity
through the transformation matrices. The stiffener element con-
sists of three nodes situated along its centroidal axis. Each node is
assumed to have six degrees of freedom, three translation u, v, w
in the directions of global axis X, Y, and Z respectively and three
rotation a, b, g about the three local axis as shown in Fig. 10.16.
The steps followed in the derivation of the element stiffness
matrix for the stiffeners are similar to eight noded shell/plate
element. The shape function for the three noded stiffener element is
given by the following equations
N1 =
1
x (x - 1)
U|
2
N 2 = (1 - x )2
|V (10.41)
1
N 3 = x(1 + x )
||
2 W
where the centroidal axis of stiffener lies along x direction [Refer
Fig. 10.16(a) and (b)].
The displacements are expressed using the above shape
functions.
2
1
3
(a) x
V3
Z
g
V2
a
V1
b
1
2 z
h
3
W x
Y
V
U X
(b)
Figure 10.16 (a) One-dimensional stiffener element.

(b) Three-dimensional stiffener element.
10.2.11 Assembly and Solution

To analyze any structure (and tank in particular), the element
stiffness matrix of all the elements in the structure are assembled.
Where structural symmetry exists as in the case of a tank, only
the symmetrical-half needs to be analyzed. Assembly of element
stiffness matrix is possible by transformation of these matrices with
reference to a common system of coordinates known as global
system. The boundary conditions for the structure provide the fixed
values of displacements in specified directions. The reactions at the
supports and the loads at given points are other input data. From
these set of known values, the displacement/loads are computed at
other remaining nodes of the structure. This results in a very large
number of simultaneous linear equations to be solved. Unless a
computer with very large core memory is available, one needs
special computer program for solution of such a large number of
equations. One particular solution package using line storage on
disc/tape is known as frontal and found to be suitable for large-size
structural problem.
10.2.12 Salient Features of Computer Program

A generalized computer program based on FEM (finite element
method) has been developed to analyze in-plane, bending and sheer
stresses at all the nodal points of transformer tanks of any kind of
profile and structure in particular. This package can also analyze,
in general, any plate/sheet structures with or without stiffeners.
The computation also accounts for the thickness of plates/shells. It
also caters for various types of loading applicable on tanks, such as
oil pressure, vacuum, gravity loads, point loads, dynamic loads etc.
The basic blocks of computation are illustrated in the flow chart
given in Fig. 10.17.
10.2.13 Results
A typical tank for a large transformer is shown in Fig. 10.18. The
element node relationship for side wall is shown in Fig. 10.19. Due
to symmetry of tank, only half of the tank has been taken. The
results for normal deflections under vacuum pressure are shown in
Fig. 10.20. The contours for major principal stresses at top under
vacuum pressure are plotted in Fig. 10.21. The computer output
provides the information about unsafe elements as well as weld
sizes for stiffener.
Start
Input basic data on

tank and loadings
Automatic generation of
coordinates of tank,
element node relations, etc.
Input data of irregular features
Is it for final run
No Yes
Print out of input data tank Compute ELT stiffness matrix

nodal coord. node
relationship, etc., for checking
Assembly of stiffness matrix
Stop
Solution of EQN
Print out of stresses

displacements, unsafe
ELTs, etc.
Stop
Figure 10.17 Flow chart.
10.3 Testing of Tanks
The oil pressure and vacuum testings are conducted on tanks to en-
sure against leakages and to check for strength.
10.3.1 Oil Pressure Test

The oil pressure testing on tank is done to ensure leakage proof
welding joints. For this the oil connection is made at the base of tank
and all the openings are blanked properly. The oil is filled up to tank
Figure 10.18
cover and the required pressure is applied by a pump. The pressure

is maintained for a few hours and all the weldings are checked
manually. If any leakage is found, the oil is drained and the welding
is rectified. After rectification oil pressure is again applied. The
tank deflection readings are measured before starting of oil
pressure, at full oil pressure and after releasing the oil pressure.
10.3.2 Vacuum Test

After oil pressure test, the oil pipe connection is opened and com-
plete oil is drained. This opening is blanked and vacuum pump con-
nection is made at top of the tank wall and oil pressure gauge is
replaced by a vacuum gauge. After ensuring all the fitting, the
vacuum pump is started and the required vacuum is measured by a
vacuum gauge. During vacuum testing the air leakage points are
detected by air leakage detecting instrument. Wherever the leak-
ages are found in gasketed rims, the bolts are tightened. The tempo-
rary deflection at full vacuum is measured. If it is considerably high
and consequently a rectification is needed, the vacuum is released
and rectification covered on the tank. Subsequently, it is again
vacuum tested. The full vacuum is maintained for one hour. After
releasing vacuum the deflection readings are taken for finding out
the permanent deflection. This shall be within specified allowable
limits of deflection, depending on the size of the tanks.
Element-node relationship
56 89 122 155 188 221 254
23 220 287
22 22 88 62 121 154 107 140 286
21 106
21 285
55 87 61 153 187 219 253 139
105 138
20 86 60 152 218 284
20
54
99
54 85 120 151 186 252

19 19 283
55 104 217 137
53
18 103
58 136
84 150 216 282
18
98
17 53 83 119 149 185 251

17 281
57 102 215 135
52
16 82 56 97 101 214 134 280
16 148
52 81
118 184 250
147 213
15 51 279
96 100
15 55 133
14 80 146 212 278
51 79 117 145 183 211 249 277
Figure 10.19 Side wall of a tank.
10.3.3 Measurement of Stresses

The strain measurement on tanks are normally not done, however,
for comparing calculated stresses with measured stresses this can
be done. The strain gauges are fixed to the tank structure with a
suitable adhesive at various locations where the stresses are
required to be measured and properly cured. The locations of strain
gauges are shown in one of the tanks in Fig. 10.22. A gauge consists
of a fine wire suitably fixed to the body of the structure. Under load
1.3 1.2 1.1 1.0 0.9 0.8 0.6
1.3
1.0
0.8
0.6
0.4 0.5
0.2
0.0
–0.1 0.4
–0.2
–0.4
–0.6 0.5
0.6
–0.7 0.7
0.3
0.8
0.2
–0.8 0.2
–0.7
–0.6
–0.5
–0.4
–0.2 0.6 0.3
0.0 0.4
0.4
0.7
0.7 0.6 0.5 0.5 0.5
Figure 10.20 Normal displacement under vacuum in mm.
strains are developed on the body, resulting in displacements of the

points to which the ends of gauges are affixed. This results in
change in the resistance of the gauge wire, which can be measured
electrically by a suitable electric/electronic bridge. One typical
strain gauge used in the experiment is a rossete-delta with six wires
connected to the respective ends of wire gauges and brought to the
bridge for measurement. The bridge is balanced for the null point
before commencement of measurements. The readings of strain
gauges are recorded simultaneously. Figure 10.23 (Plate 4) shows
18 17 17 15 14 13 12 11 10 8 6
18 6
19.2 8
–3.5 10
–25 –2.5
–3 –4
–2
–1.5
–4
–3.5
–3
–1.5
–2
–4.5
–3
–2.5 –3.5
–4
–5.0
–4
–5.5
3 –3.5
–3 –3
–2.5
–4 –2–1
–0.0
–4 –1.5
2
–3
–2.5
–2
–1.5
–1
–1 –1.5 –2 –2.5 –2 –1.5 –1.0 –1.5 2

–2.5 0.0
–1.0
Figure 10.21 Principle stresses under vacuum in mm.
the wiring connections and the instrumentation for the measure-

ment. The tank is subjected to full vacuum and readings are taken.
The strain gauges are fixed inside and outside the tank wall on
large tanks to compare the top and bottom principal stresses on the
surface [Refer Fig. 10.24 (Plate 4)].
The strain gauge gives values of strains in the direction of rossets.
From this, the two principal stresses s1, s2 and their directions may
be computed from the following expressions.
8
9
2 2
7 6 5
3 3 2 1 3
Location of strain gauge
Figure 10.22 Transformer tank showing strain gauges.
2
s1, 2 = E
e1 + e2 + e3 1
±
FG e 1 -
e1 + e2 + e3 IJ + FG e 2 - e3 IJ (10.42)
3(1 + m ) m H 3 K H 3 K
( e 1 - e 3 )/ 3
Q = tan–1 (10.43)
FG e + e2 + e3 IJ
e1 - 1
H 3 K
where E is Young’s modulus in kg/cm2
e1, e2, e3 are strains in three directions
m is the poisson’s ratio.
Nomenclature
p = Pressure in kg/cm2
b = Width of the plate in cm
l = Longitudinal length of plate in cm
t = Thickness of the plate in cm
E = Young’s modulus in kg/cm2
I = Moment of inertia in cm4
Z = Section modulus in cm3
f = Stress developed in kg/cm2
d = Deflection in cm
fss = Stresses for simply supported plate in kg/cm2
fEB = Stresses for fixed and plate in kg/cm2

IXX = Moment of inertia along X-X direction cm4
IYY = Moment of inertia along Y-Y direction cm4
K1, K2, K3 = Factors depending on boundary conditions
M = Bending moment, kg cm
e1, e2, e3 = Strains in three directions, mm
s1, s2 = Principal stresses in kg/cm2
sx = Stress in x direction in kg/cm2
m = Poisson’s ratio
Subscripts p for plate
s for stiffener
ss for simply supported
EB encasterated beam
REFERENCES
1. Timoshenko, S. and S. Woinowsky Krieger, “Theory of Plates and
Shells”, 2nd Edition, McGraw-Hill, New York, 1959.
2. Shakya, M.K., “Stress Analysis of Transformer Tanks by Strain
Gauges”, M. Tech. Dissertation, M.A.C.T., Bhopal, 1980.
3. Perry, C.C. and H.R. Lissner “Strainguage Primer”, McGraw-Hill
Book Company, New York, 1962.
4. Cook, R.D., Concepts and Applications of Finite Element Analysis,
John Wiley & Sons, Inc. 1974.
C HAPTER 11
Transformer Auxiliaries and

Oil Preservation Systems
S.C. Verma
J.S. Kuntia
Transformer auxiliaries play a vital role in ensuring proper function-

ing of the main equipment. Some of the auxiliaries provide
protection under fault conditions.
Transformer oil being a major insulation requires special attention
against contamination by moisture and oxygen for preservation of
quality.
TRANSFORMER AUXILIARIES
11.1 Gas Operated (Buchholz) Relay
The relay serves as main protection for any minor or major faults
that may develop inside a transformer. Such faults always result in
generation of gases which causes the operation of mercury switches
giving signal for audible alarm or isolates the transformer from the
network.
Figure 11.1 (Plate 4) shows the internal view of a relay. It
comprises a cast housing which contains two pivoted buckets, each
bucket being counter-balanced by a weight. Each assembly carries
a mercury switch, the leads from which are taken to a moulded
terminal block.
Operation
The relay is mounted in the pipe at an inclination of 3–7° as shown in
Fig. 11.2. In healthy condition of the transformer, the relay is full of
3 – 7°
1. Tank
2. Conservator
3. Buchholz-relay
Figure 11.2 Method of mounting Buchholz-relay on transformer.
oil and both the mercury switches are open. In the event of a minor
fault like damage to core bolt insulation, local overheating, etc., the
arcing causes slow generation of gas in the oil, which passes up in
the pipe and gets trapped in the relay housing. As gas accumulates,
the oil level in the relay falls, leaving the top bucket full of oil.
When a sufficient volume of gas is collected in the relay, the
top bucket, because of its extra weight due to oil contained, tilts,
overcoming the balance weight which closes the mercury switch and
initiates an audible alarm.
With a major fault like short-circuit between turns, coils or
between phases; the generation of gases is rapid and the gas and the
displaced oil surges through the relay and impinges on the baffle
plates, causing the lower assembly to tilt and close the mercury
switch and provide singal for tripping the circuit breaker, which
disconnects the transformer from the network.
Transformer Auxiliaries and Oil Preservation Systems 253
11.1.1 Relays for Earthquake Zones
Shock and vibrations acting on tube of the mercury switch can cause
the mercury within it to move and momentarily bridge the switch
contact, even though it may be tilted in open-circuit position. This is
considered to be a mal-operation of the relay, as it is caused by exter-
nal disturbances like earth tremors and not by a fault within the
transformer. Consequently, relays for such situations are provided
with magnet operated reed switches in place of mercury switches.
The external appearance and principle of operation of these relays
is exactly similar to those with mercury switches. The reed type
switches have rhodium contacts located midway along the glass tube
which has an atmosphere of nitrogen.
11.2 Temperature Indicators
Temperature indicators are precision instruments, specially

designed for protection of transformers and perform the following
functions.
— Indicate maximum oil temperature and maximum or hottest
spot temperature of winding.
— Operate an alarm or a trip circuit at a predetermined
tem-perature.
— Switch on the cooling equipment when the winding attains a
preset high temperature and switch it off when the tempera-
ture drops by an established differential (so as to avoid too
frequent on and off operation of the switch).
Normally two separate instruments are used for indicating oil and
winding temperatures.
11.2.1 Construction and Principle of Operation
These indicators normally work on the principle of liquid expansion,

the liquid being sealed in the bellows. Figure 11.3 shows a sectional
view. The indicator is provided with a sensing bulb placed in an oil
filled pocket on the transformer tank cover. The bulb is connected
to the instrument housing by means of flexible connecting tubing
Capillary
Bulb
Compensating lever Anchor post

Operating bellows
Bellows heater coil Compensating bellows
Range adjuster
Rack and pinion movement
Movement arm
Potentiometer winding
Side link
Screws
Terminal block
Switch zeroing screw Switch setting screws
Switch platform
Figure 11.3 Sectional view of a temperature indicator.
consisting of two capillary tubes. One capillary is connected to the

operating bellow of the instrument and the other to a compensating
bellow. The operating system is filled with a liquid which changes
volume with varying temperature. The compensating bellow
acts upon the operating bellow through a linkage compensating for
variations in the ambient temperature. With change in the volume
of the liquid, the bellows expand or contract, transmitting the
movement through a linkage mechanism to the indicating pointer
and switching disc. Up to four dry electrode mercury switches, each
mounted on a steel carriage, are provided on a winding temperature
indicator. The make-and-break temperature of each switch can be
independently adjusted.
Oil and winding temperature indicators work on the same
principle, except that winding temperature indicator is provided
with an additional bellow heating element. As it is not possible to
measure the winding temperature directly, it is done indirectly by
means of a thermal image process.
The heating element is fed by a current transformer, with a
current proportional to the load in the winding temperature is to be
indicated. The temperature increase of the heating element is
thereby proportional to the temperature increase of the winding
over top oil temperature. As bulb of the instrument is located in

the hottest oil zone, it senses the maximum oil temperature. The
operating bellow thus gets additional movement, simulating the
increment of winding temperature above maximum oil temperature
and thus indicates hottest spot temperature of winding.
11.3 Pressure Relief Valve
The pressure relief valve plays a significant role in the protection of

power transformer systems. As mentioned before, a major fault
inside the transformer causes instantaneous vaporization of the oil,
leading to extremely rapid build-up of gaseous pressure. If this
pressure is not relieved within a few milliseconds, the transformer
tank can get ruptured, spilling oil over a wide area. The consequent
damage and fire hazard possibilities are obvious. A pressure relief
device provides instantaneous relieving of dangerous pressure.
11.3.1 Construction and Operation
Figure 11.4 shows a cross-sectional view of the valve. The valve is

generally mounted on the tank cover above an opening. The valve
has a corresponding port which is normally sealed by a stainless
steel diaphragm (4). The diaphragm rests on an ‘O’ ring (3) and is
kept pressed by two heavy duty springs (6). As the pressure inside
the tank rises above a preset limit due to a major fault, the
diaphragm gets lifted instantaneously and excessive pressure drops,
the diaphragm then restores to its original position. The lift of
the diaphragm is utilized to operate a flag indicator (10) and a
micro-switch with the help of a rod (8). Figure 11.5 (Plate 4) shows
the photograph of the spring loaded type of pressure relief device.
Another type of device which is also used for the same purpose is
called explosion vent. Figure 11.6 shows the general construction of
an explosion vent. In the event of a serious fault, due to excessive
pressure, the top diaphragm ruptures, thus releasing the pressure.
Due to certain superior features of spring loaded pressure relief
valve like smaller size, elimination of equalizer pipe and provision of
a switch for alarm annunciation in event of its operation, it is finding
a wide spread preference over the explosion vent.
10
9
8 1. Base
2. Gasket ring
3. O’ ring
7 4. Diaphragm
5. Cover
6 6. Spring
5 7. Rod retaining spring
8. Switch operating rod
4 9. Flag carrier plate
10. Flag indicator
3
2
1
Figure 11.4 Sectional view of a spring-loaded pressure relief valve.
11.4 Oil Level Indicator
Normally all transformers are provided with an expansion vessel

called conservator, to take care of expansion in the oil volume due to
rise in temperature, when the load on the transformer increases
or due to increase in ambient temperature. The oil level in the
conservator consequently goes up. Conversely, it falls when the
temperature or load reduces. It is essential that the oil level in the
conservator is maintained above a pre-determined minimum level.
All large transformers are, therefore, fitted with a magnetic oil level
gauge which also incorporates a mercury switch. The switch closes
and actuates an audible alarm in the event of oil level dropping to
near empty position in the conservator. Figure 11.7 (Plate 5) shows
the photograph of such an indicator.
11.4.1 Construction of Operation
Figure 11.8 shows a cross-sectional view of the inner details and a

schematic diagram. A float is used as a sensor which moves with the
1. Equalizing pipe
2. Oil inlet hole
3. Deflector diaphragm
4. Slot for air release
4 3
Figure 11.6 Explosion vent.
rise and fall of the oil bevel. Its movement gets transmitted to the
switch mechanism by means of bevel gear and magnetic coupling,
which ensures a complete seal between the conservator and switch
compartment. The pointer is also magnetically operated and picks
up the correct oil level.
11.5 Bushing and Cable Sealing Box
It is necessary to bring the low and high voltage leads out of trans-
former tank, to be able to make connections between transformer
and generator or transmission lines, etc. This is accomplished by
terminating these leads through what are known as bushings or
cable box.
A bushing is a structure carrying a conductor through a partition
in the tank and insulating the conductor from partition.
11.5.1 Cable Sealing Box
Generally used for termination of leads of low voltage, a cable

sealing box is designed for the purpose of receiving and protecting
7 6 5
1. Float
2. Float arm
4
3. Bevel gear
3 4. Limit stoppers
5. Main body
6. Switch housing
2 7. Mercury switch
1 8. Dial body
9. Driving magnet (for pointer)
10. Cam
11. Driving magnet (main coupling)
12. Driven magnet (main coupling)
13. Magnet with pointer
8 9 10
Schematic
Dial
10
3
2
Pointer 12 11 1
Air Oil
13 9
Figure 11.8 Sectional view and schematic of a magnetic oil level indicator.
the end of a metal-sheathed cable or cables and containing a suitable

insulating medium.
It is a unit complete with bushings, to which the terminals of the
transformer can be connected. The insulating medium in a cable box
can be air or a bituminous compound. Figure 11.10 shows a typical
3-pole, 9-gland cable sealing box.
TRANSFORMER OIL PRESERVATION SYSTEMS
Transformer oil deterioration takes place due to moisture. Moisture

can appear in a transformer from three sources, viz. by leakage past
gasket, by absorption from air in contact with the surface, or by its
formation within the transformer as a product of deterioration as
2
9
7
1
6
8
5
10
1. Bushing 6. Flexible connection

2. Shell 7. Connector
Transformer Auxiliaries and Oil Preservation Systems
3. Cable gland 8. Cable ferrule

4. Cable (insulated) 9. Opening
5. Cable (bare) 10. Filling material (insulation)
259
Figure 11.10 A three-pole, nine-gland cable box.

insulation ages at high temperature. The effect of moisture in oil is

to reduce the electric strength, especially if loose fibres or duct
particles are present.
Method available to reduce oil contamination from moisture are
silicagel breather, thermosyphon filter, sealed conservator tanks
using gas cushion, rubber diaphragm or air-cell seals refrigerated
dryers.
11.6 Silicagel Breather
A silicagel breather is most commonly employed as a means of

preventing moisture ingress. It is connected to the conservator tank,
which is fitted to transformer to allow for changes in volume due to
temperature variations. As the load reduces, air is drawn into the
conservator through a cartridge packed with silicagel dessicant,
which effectively dries the air. Freshly regenerated gel is very
efficient, it will dry the air down to a dew point of below –40°C, but
quickly falls in efficiency. A well maintained silicagel breather will
generally operate with a dew point of –35°C, as long as a large
enough quantity of gel has been used for the cycling duty. Figure
11.11 shows such a breather.
Silicagel may be reactivated by heating in a shallow pan at a tem-
perature of 150° to 200°C for two to three hours when the crystal
should have regained their blue tint.
11.7 Gas Sealed Conservators
In this method, the contact between transformer oil and atmospheric

air is eliminated by providing cushion of an inert gas over oil surface
in the conservator vessel. The gas pressure is always higher than
atmospheric pressure to avoid ingress of air. The gas normally used
for this purpose is nitrogen having high purity and dryness.
Construction and Operation

Figures 11.12 (Plate 5) and 11.13 show respectively, a photograph
and a schematic diagram of nitrogen sealing system. High pressure
nitrogen gas at 15 MPa flows out of the cylinder and is admitted in
Connecting pipe
to transformer
Filling hole
Container
When the
silicagel
becomes pink
regenerator
or change
the silicagel
View glass
Silicagel filling
Strainer
Drain hole
Air inlet
Oil seal
Figure 11.11 Silicagel breather.
the conservator after passing through multi-stage pressure reducing

valves. The pressure reducing valves automatically cut off the
nitrogen gas supply when the pressure in the conservator reaches
3–5 kPa. Due to increase in ambient temperature and load the gas
pressure builds up.
The system is designed to relieve any excessive pressure through
a relief valve provided for the purpose. When the pressure drops
below 3–5 kPa, the valves open to admit nitrogen from cylinder and
this cycle continues until the cylinder becomes empty.
To conser-
vator
Isolating
PVC tube valve
connected + + HP gauge with
to atmosphere one contact
Multi-stage
+ + regulator
10
Isolating
valve
+ +
LP gauge with
two contacts
N2 cylinder N2 cylinder
+ +
+ +
+ 21
+ +
Isolating
valve
+ +
Pressure
relief valve
Oil pump Cabinet
Figure 11.13 A typical general arrangement of nitrogen system.
The system may be provided with switches in the pressure

gauges to operate alarm under the following abnormal operational
conditions:
(a) When the pressure in transformer exceeds relief over valve
pre-set operating pressure.
(b) When the pressure in transformer drops below 3–5 kPa.
(c) When the cylinder pressure is reduced to around 1 MPa.
11.8 Thermosyphon Filters
Thermosyphon filters are intended for prolonging transformer oil

life by extracting harmful constituents like water, acids, etc., from
oil. Such filters are normally installed on ONAN or ONAF cooled
transformers.
The constructional details of such a filter are shown in Fig.11.14.
Normally the filter is mounted directly on the transformer tank. The
filter, generally of cylindrical shape, has a number of perforated steel
trays filled with an adsorbent material. The adsorbents generally
employed are silicagel and active alumina, the latter being slightly
more effective.
1
1. Shell
2. Trays containing alumina
2
Figure 11.14 Thermosyphon filter.

As a result of difference in the temperature between the upper

and lower layers of oil in the tank of a transformer in operation, the
oil circulates through the filter by convection currents. The water
absorption capacity of alumina at relative humidity of 60% is about
15% by weight.
11.9 Bellows and Diaphragm Sealed Conservators
The contact between atmospheric air and transformer is prevented

by a barrier which is made from a synthetic rubber compound. The
general principle of such a system is shown in Fig. 11.15. In case of
bellow type of barrier, as the oil level in the conservator vessel falls,
air is sucked from the atmosphere through a silicagel breather,
inflating the bellow. The bellow deflates as the oil level goes up. The
conservator is also fitted with pressure vacuum bleeder to pass
either oil or such in air in the event of over-filling or under-filling of
conservator.
Bellow Hand hole

Pressure vacuum valve
Drain valve
Breather
Figure 11.15 Bellow sealed conservator.
When diaphragm is used as a barrier between oil and atmospheric

air, the conservator vessel is made in two semi-circular halves as
shown in Fig. 11.16. The diaphragm is held between the two halves
Diaphragm
Oil level
gauge
Drain valve
Breather
Figure 11.16 Diaphragm sealed conservator.
and bolted. As oil expands and moves up, it pushes the diaphragm
upwards. The position of the diaphragm is indicated by the oil level
indicator as the oil level indicator connecting rod is connected to the
diaphragm. When the oil level falls down in the conservator, the
diaphragm deflates creating vacuum, which is filled by air getting
sucked through a silicagel breather.
These type of sealing systems have one advantage over the
gas-sealed conservator. If gas is pressurized to a high level, it gets
dissolved in oil. Over a period of time the amount of gas in oil
reaches the saturation point. If at this stage, the load on transformer
is suddenly dropped, or the ambient temperature falls severely, the
pressure falls, oil becomes supersaturated and gas bubbles will be
evolved. If there is a pump connected in the cooling circuit, it will
help generation of bubbles. These bubbles may cause insulation
failure in the region of strong electric fields.
11.10 Refrigeration Breathers
In the refrigeration breather system, an air dryer is fitted to the

conservator vessel. Air breathed through the unit is dried in passing
down a duct cooled by a series of thermoelectric modules based
on Peltier effect. Top and bottom ends of the duct are terminated
in the expansion space above oil level in the conservator and air
continuously circulated through the duct by thermosyphon forces.
Figure 11.17 shows general constructional feature of a refrigera-

tion breather system. When transformer is breathing, warm moist
air passes through the central duct; its cold surface which is held at
about – 20°C, collects any water vapours in air and condenses them
as frost and ice. Periodically, the frost and ice on the duct is removed
by reversing the current to the thermoelectric modules. The current
reversal heats up the duct walls sufficiently to melt the ice, and
passes out as water through a small drain tube. The current flow is
again reversed in the modules and the cycle repeats automatically by
a control unit.
5 6 1
1. Conservator 4.Thermoelectric modules

2. Warm and moist air 5. Air inlet tube
3. Central duct 6. Drain tube
Figure 11.17 Refrigerated breather system.
Atmospheric air is breathed in through a small tube entering the

unit at the top and has to pass through the central duct, ensuring
that it reaches the conservator in dry state.
C HAPTER 12
Manufacturing and Assembly

T.K. Ganguli
M.V. Prabhakar
Manufacturing and assembly of the transformer involves the follow-

ing main stages and the manufacturing techniques basically depend
on the design philosophy.
(a) Core building
(b) Preparation of windings
(c) Core and winding assembly
(d) Terminal gear assembly
(e) Processing
(f) Servicing and tanking
(g) Mounting of accessories for testing
12.1 Core Building
The core forms the magnetic circuit of the transformer. Core is built
up from cold rolled grain oriented, silicon alloy sheet steel of the best
magnetic properties. The material is either received in rolled coil
form or in finished sizes of core sheets, ready for core building. When
the material is received in coil form, it is slitted, cropped and mitred
to the required dimensions based on the design requirements. CNC
machines are now available which can perform the above operations
with perfect control on burrs. The size, type, and construction of the
core, like the number of steps, number of limbs, height, etc. depend
on the number of phases and the size of the transformer.
The sequence of operations for core building and assembly are as

under:
(a) Slitting of core steel rolls to required width on slitting
machines.
(b) Cropping and mitring to the required dimensions.
(c) Hole punching in the laminations where required.
(d) Stacking of laminations of different size to the required
thickness.
(e) Laying of clamp plates and end frame and its leveling.
(f) Assembly of insulation between clamp plate/end frame to core
laminations.
(g) Preparation of oil duct in core.
(h) Core building.
(i) Clamping of core after assembly of the top end frame.
(j) Tightening of core.
(k) Lifting of core by use of a cradle, and carrying out isolation
checks after treatment of insulation items.
The core assembly is now ready for further processing and
assembly.
12.2 Preparation of Windings
The windings form the electrical circuit of a transformer. Various

types of windings, i.e. spiral, helical, disc, layer, etc. are used,
depending upon the rating of the transformer and the design consid-
erations. Generally, low voltage windings are either spiral or helical
and high voltage windings are either layer or disc type. Various types
of windings have been discussed in detail in Chapter 5.
The sequence of manufacture of windings is given below:
(a) Loading of moulds (formers) on the winding machines.
(b) Loading of the conductor reels on stands.
(c) Dressing of the mould, i.e. assembly of insulation spacers and
blocks on the mould.
(d) Manufacture of the winding on horizontal/vertical winding
machines depending on design, type of winding, number of
conductors, to be handled at a time and the type of conductor.
(e) Preparation of the leads etc.
Manufacturing and Assembly 269
(f) Dismantling of winding from the machine.

(g) Preparation of the joints between conductors, if any.
(h) After removing the winding from the winding machine, each
winding is clamped between top and bottom plates through tie
rods, and kept in an oven for heating. The windings are
individually shrunk to the required axial dimensions by
heating in the steam heated oven and by applying the
required pressure. Heating ensures removal of moisture from
the insulation items. This process is called stabilization of the
windings, and the windings are stabilized to such an extent
that they do not shrink further during service.
12.2.1 Layer Winding Assembly
Layer winding is also clamped and shrunk as above. Winding is then

placed with bottom end in top position and the bottom end insulation
consisting of angle rings, angle washers, flanged paper insulation,
block washers, etc., is built up. The position of winding is then
reversed, bringing top of winding on top and the top insulation
arrangement is completed in this position. During this process,
insertion of line shield rings on both ends and filler paper, prepara-
tion of leads, placement of circular barriers, etc., are also completed.
After completion of insulation arrangement, the coil is shrunk to the
required axial dimensions.
Typical insulation arrangement of layer winding is shown in
Fig. 12.1.
The winding is then kept in clamped condition for either compos-
ite winding assembly or assembly on core legs.
12.3 Winding Assembly
A typical arrangement of winding assembly along with clamping

structure is shown in Fig. 12.2 (Plate 5).
The winding assembly is carried out in clamping structure
consisting of bottom and top face plates and tie rods. Steel face plates
are selected on the basis of maximum diameter of the winding to
be assembled. Number of windings to be assembled depends upon
the type of transformer, i.e. whether generator transformer auto-
transformer or system transformer, etc. Generally, in a generator
1. Shield ring 4. Angle washer

2. Filler paper 5. Block washer
3. Angle ring 6. Barrier
Figure 12.1 Typical insulation arrangement of layer winding.
transformer low voltage, tapping and high voltage windings are

involved whereas in auto-transformer tertiary, common, tapping
and series windings are involved. A typical sequence of operation for
winding assembly of a transformer having three windings when they
are pre-assembled as follows:

(a) Winding 2 is lowered over winding 1, placing radial spacers
and pressboard cylinders in between and keeping bottom
end of each coil at the top. Next winding 3 is lowered over
winding 2.
(b) Insulation items, i.e. block washers, angle rings, etc., are
assembled. During insulation assembly, various leads i.e.
tapping, line leads, etc., are properly positioned and partly
insulated at this stage.
(c) Winding assembly is reversed to bring top of the coil at the
top. Insulation items of top portion are then assembled, and
various leads are also prepared.
(d) Barrier arrangement is then assembled and tied with tape bands.
(e) Winding assembly is clamped in clamping structure and is
shrunk to the required axial dimensions. The assembly is
kept under pressure for further assembly on core legs.
The individual windings alternatively can also be assembled on
the core leg instead of being pre-assembled.
12.4 Core and Winding Assembly
Core is placed on suitable levelled platform. Top-end frame and yoke

laminations are removed. Bottom insulation items are placed on bot-
tom yoke/end frame on each core leg. Various windings are either
lowered one by one or composite winding assembly is lowered on
each core leg. After lowering of the windings, top insulation arrange-
ment is completed. Top-end frame and yoke laminations are placed
back in position. Coils are kept under pressure by either coil clamp-
ing bolts or spring loaded hydraulic devices provided on the top-end
frame. A typical arrangement of core and winding assembly is shown
in Fig. 12.3 (Plate 5).
12.5 Terminal Gear Assembly
After relacing of the top yoke, the preparation of the Terminal Gear
Assembly is done as described below:
(a) Cutting of the leads as required.
(b) Crimping/brazing of the leads with cables.

(c) Brazing of bus bars.
(d) Fixing of different cleats.
(e) Crimping/brazing of cables with terminal lugs.
(f) Mounting of the tap changer/tap switch.
(g) Preparation of HV line lead.
In this stage, connections between phases to form the required
vector group, tapping lead connections, line and neutral leads forma-
tion, etc., are completed. Low-voltage connections are done on one
side of the winding and are designated as LV terminal gear. On the
opposite side, high-voltage connections are done and are designated as
HV terminal gear. Medium voltage leads (in system or auto-trans-
former) are taken out on LV side and tapping connections on either
LV or HV side depending upon design layout. Generally in generator
transformer, a three-phase on-load or off-circuit tap-changer is
mounted on one end and in case of auto-transformer three single-
phase tapchangers are mounted in front of the windings. Tapchangers
are supported from end frame during terminal gear assembly.
All leads, i.e. line and neutral leads of low-voltage, medium-
voltage and high-voltage windings, tapping leads, etc., are laid out
and connected using different types of joints (i.e. bolted, crimped,
soldered or brazed) and insulated for the required insulation level.
Leads are properly supported by cleats mounted on end frames.
The clearances between various leads, coil to leads, leads to end
frame and other parts are maintained and checked. Figures 12.4(a)
and (b) (Plate 5) show photographs of typical LV terminal gear
arrangements.
Figures 12.5(a) and (b) (Plate 6) shows photographs of typical HV
terminal gear arrangements.
12.6 Placement of Core and Winding Assembly in Tank
After completion of terminal gear assembly, the core and winding

assembly is placed in the tank. The tank may either be of conven-
tional or bell-shaped construction depending upon requirement. In
case of bell-shaped construction, the top part of tank and for conven-
tional construction the bottom tank is prepared for this purpose.
Preparations include mounting of shunts, barriers, etc., on tank

walls as required and also laying of gaskets on the flange joints. In
conventional tank, the core and winding assembly is lowered in the
bottom tank. [Fig. 12.6(a) (Plate 6)]
In bell-shaped tank construction, top tank is lowered on the core
and winding assembly. [Fig. 12.6(b) (Plate 6)]
After placement of core and winding assembly in tank, various
electrical and mechanical clearances, viz. coils to tank, line, neutral
and tapping leads to tank, etc., are checked. After the above checks
are over, all the openings in tank and cover are blanked by blanking
plates and gaskets. The transformer is then sent for drying and
impregnation.
12.7 Processing
The quality of a transformer and, consequently its performance and

life depends essentially on the factory processing. The assembled
active part of the transformer after tanking or before tanking as the
case may be, is heated in an oven for extraction of moisture under
vacuum. Sometimes, vacuum drying is done by drawing vacuum in
its own tank and keeping the whole transformer in an oven.
Besides the conventional method of drying, the most modern
vapour phase drying method is also used, as explained in detail in
Chapter 13.
Much importance is attached to proper dry out, as otherwise wind-
ings would become loose in service, and may get distorted under
short circuit conditions.
Once the moisture is extracted to the desired level, oil is filled
under vacuum and the transformer is soaked in oil.
12.8 Servicing of Transformer
The transformer active part is then taken out and all the terminal
gear cleating of the leads are retightened and the coils are clamped
and secured in position. The transformer tank cover is then placed in
its position and the transformer is now ready for further assembly of
bushings etc.
12.9 Tanking
After servicing the transformer is assembled with all the necessary

fittings. The oil is filled under vacuum and it is circulated in order to
get the desired ppm and BDV levels before being offered for final
testing and subjected to high voltage testing.
C HAPTER 13
Drying and Impregnation

M.P. Singh
M.V. Prabhakar
In the construction of a transformer, the insulation system is the

most important feature, and hence requires maximum attention.
Normally, the oil, paper, and pressboard insulation has to be freed
of dirt, dust, and moisture for obtaining the optimum insulating
properties.
Cellulose insulation used in power transformers and reactors has
8 to 10% of moisture by weight at ambient temperature, being a
hygroscopic material. Water is injurious to transformer insulation
system, since it reduces the electric strength and the resistivity of
the oil and accelerate the deterioration of solid insulation. Figure
13.1 shows the effect of moisture content on impulse voltage
strength of oil and paper. The decrease in impulse voltage strength
is clearly evident as soon as the moisture content of the paper is
greater than 0.1%. Paper with a 1.5% moisture content ages ten
times faster than on with only 0.3% as shown in Fig. 13.2. Processing
eliminates all types of impurities and foreign particles from trans-
Impulse strength %
100
80
60
40
20
0 2 4 6 8 10 12 14 16 18 20ppm
Water content in oil
0.1 0.15 0.2%

Water content in paper
Figure 13.1 Impulse voltage strength of paper.

100
Multiple of ageing time 10
1
0.1 1 Weight % 10
Water content X
Figure 13.2 Influence of moisture in paper on ageing time of paper:

a multiple of the ageing time of oil-impregnated paper
with moisture content X compared to that of a well-
dried, oil-impregnated paper with 0.3% residual
moisture.
former, e.g. dust, dirt, fine metallic particles and fibres of diverse
origins. Any impurity and particle which remains inside transformer
after vacuum drying is taken away by transformer oil during circula-
tion, which are separated at the filter. During vacuum drying and oil
impregnation, the gases trapped inside solid insulation and occluded
among intricate profile of insulation components are also removed,
which result in low partial discharge inception voltage. Optimal
utilization of the excellent properties of oil-cellulose dielectric
system depends upon efficient drying, degassing and impregnation
process adopted and hence the proper processing of transformers is a
vital requirement.
13.1 Basic Principles of Drying
Cellulosic insulation is dried by creating conditions in which water

vapour pressure (WVP) around the insulation is less than that in
the insulation. The vapour pressure in the insulation is increased
by heating the insulation and the vapour pressure around the
insulation is decreased by removing water vapour. Figure 13.3
shows pressure plotted against humidity and temperature, from
which it is seen that a 20°C temperature rise increases the internal
pressure by more than 100% (a factor of two). Basically one should
aim at achieving the highest processing temperature consistent with
Drying and Impregnation 277
8 5%
1000
2,5
1
100 0,5
H2O-Partial pressure torr
0,3
10 0,1
0.1
0 20 40 60 80 100 120 140
Temperature°C
Figure 13.3 Partial insulation water vapour pressure v. temperature,

for various insulation moisture contents.
the type and ageing properties of the insulation. The upper tempera-
ture limit is set by the maximum permissible drying temperature for
paper, i.e. about 110°C (or 130°C in an oxygen-free atmosphere). Dry-
ing efficiency also depends on diffusion coefficient of the insulation
material. This coefficient is dependent on the material to be dried,
its temperature, pressure and moisture content as shown in Fig.
13.4. A 20°C temperature rise increases the diffusion coefficient by
about 100% and reduces drying time by half.
Two basic processes are adopted for drying of transformers.
(a) Conventional vacuum drying
(b) Vapour phase drying
PRINCIPLES OF DRYING
13.1.1 Drying Time
The drying time of the insulation increases considerably with the

operating voltage and size of the transformer. The longer the drying
time, the higher the partial discharge inception voltage.
°C °C
30 120 300 120
110 110
100
20 200
Coefficient of diffusion cm2/day
Coefficient of diffusion cm2/day

100 90
80
90
10 100 70
80
60
70
5 50
50
3 60 30
2 20
50
1 10
0.5 5
0 2.5 5 7.5 10 0 2.5 5 7.5 10
Humidity % Humidity %
(a) (b)
Figure 13.4 Diffusion coefficient of low density oil-free press-

board for various drying temperatures.
(a) under atmospheric pressure.
(b) over a 0.1 to 1.0 torr vacuum range.
The time of drying depends on the final moisture content required
in the insulation pressure rings, supporting components and pre-com-
pressed press boards basically determine the minimum drying time.
13.2 Conventional Vacuum Drying
The moisture content in a cellulose insulation can be brought down

to 2% from 8–10% by dry hot air circulation. For further reduction of
moisture content, heating under vacuum is resorted to, so that the
time required for drying can be reduced.
The core and windings assembly is placed in the clean transformer
tank. Drying is carried out either in a vacuum vessel or in an air
circulating drying oven. In the latter case, vacuum is drawn in the
transformer tank.
13.2.1 Loading of Transformer
The tank containing core and windings is loaded into the vacuum
vessel. For bell shaped tanks the openings on the top of the tank are
kept open such that complete cellulose insulation is oil impregnated

subsequently. For drying in its own tank, the tank alongwith core
and windings is loaded into the drying oven. Tank cover is placed in
position and all openings are closed.
13.2.2 Vacuum Drying
The transformer is heated initially to 100°C for about 24 hrs and a

high vacuum in vessel or in its own tank is drawn. Higher the
vacuum, better and quicker is the drying. Vacuum level to be
achieved also depends upon voltage class of transformer. Vacuum
corresponding to an absolute pressure as low as 1.33 Pa (0.01 torr)
can be obtained. In order to maintain the transformer temperature
at the required value, the pressure is increased to atmospheric level
by the admission of hot dry air at intervals during the first few
days of drying. This is necessary since winding temperature drops
considerably as latent heat of transformer is utilized in conversion of
insulation moisture into water vapours.
During vacuum drying, the quantity of water collected at pump is
recorded at regular intervals. Insulation resistance, power factor
and dispersion factor of windings are also monitored. Vacuum drying
is considered to be complete, when desired condensate extraction
rate is achieved and other parameters indicated above become
reasonably steady.
13.2.3 Reconditioning of Insulating Oil
There is a tendency for oil to absorb water due to breathing during

transport and storage. Impurities, be they solid ones like hygroscopic
fibres, suspended particles or liquid ones like dissolved water, bring
about a considerable reduction of the dielectric strength of oil. Oil at
20°C saturated with water (44 ppm) attains only about 25% of the
original electric strength with a water content of 10 ppm. Under the
same conditions, the resistivity of oil also drops considerably. Air,
especially oxygen, dissolved in the oil represents a risk not only
with regard to the formation of bubbles but also due to accelerated
oxidation process leading the chemical deterioration of oil. Hence, it
is necessary to treat and purify the oil before impregnating the wind-
ings, so as to attain a degree of purity which meets the operational
requirements. Oil is reconditioned to eliminate the following ele-

ments.
(a) Solid impurities
(b) Free and dissolved water
(c) Dissolved gases
The physical means that are used for treatment of oil include
several types of filtration, centrifuging and vacuum dehydration
techniques. If vacuum treatment is employed, temperatures up to
80°C can be used; if not, it is advisable to limit the temperature of oil
to 60°C to prevent oxidation. The filters used should be capable of
removing particles larger than one micron diameter.
Along with drying of transformer, raw oil is reconditioned in a
vacuum dehydrator, till desired characteristics are obtained.
13.2.4 Oil Impregnation
After completion of vacuum drying, the reconditioned oil is admitted

to the transformer without breaking vacuum in the vessel or tank,
as the case may be. The transformer is allowed to cool to a tempera-
ture of about 70°C before filling with oil. Temperature of recondi-
tioned oil is maintained at 50 ± 10°C. Care is taken to ensure that
pressure in vessel or tank is not too low to vaporize the oil. The oil
flow rate normally does not exceed 3 kl/h so that all residual gases
occluded in crevices among insulation items are released and the
possibility of gas bubbles being trapped at any place inside tank is
prevented. Oil is filled until all cellulose insulation is covered.
13.2.5 Redrying and Oil Circulation
Oil impregnated transformer is kept under vacuum for a few

hours so that solid insulation is completely soaked with oil. Winding
insulation shrinks to some extent during drying as the initial water
in cellulose is extracted. Hence, windings and cleats of terminal gear
are retightened after taking the assembly out of tank. A little mois-
ture reabsorbed by the insulation due to exposure to atmosphere
during this period is extracted by redrying under vacuum and
subsequently oil is again impregnated inside the transformer. Oil is
circulated to regain its electric strength and moisture content which
deteriorate due to contact with materials used in the construction of
transformer. Solid particles, fibres of different origins and all other

impurities are separated at filter. A settling time is given before con-
ducting high-voltage dielectric tests on the transformer, so that sus-
pended fine particles are settled and occluded gas is fully released.
13.3 Vapour Phase Drying
The main difference between conventional vacuum drying and

vapour phase drying is that, in the latter process the heat carrier is
a vapour of low viscosity solvent, more like kerosene, with a
sufficiently high flash point instead of air. The vapour is condensed
on the transformer and then re-evaporated in the plant. For this
reason, vapour phase installations include an evaporator and
condenser system in addition to the vacuum equipment and vacuum
vessels associated with conventional drying equipment. Typical
schematic of a vapour phase drying system is shown in Fig. 13.5,
which is applicable to transformers dried in the vacuum vessel as
well as in their own tank. It may be seen from Fig. 13.5 that solvent
heat conveyor system consists of storage, evaporation, condensation
filtration, solvent feedback and control arrangement.
13.3.1 Heat Carrier
The solvent used should possess the following properties for effective
and efficient drying.
(a) Vapour pressure must be distinctly below that of water, so
that a large pressure difference assists efficient water diffusion
from the beginning of the heating phase (Fig. 13.6).
(b) Evaporation heat should be as high as possible.
(c) The presence of small amounts of the heat-carrier in the solid
or oil insulation must have no effect on their ageing or
general properties.
(d) High ageing stability, allowing practically unlimited use in
the drying process. A solvent storage tank is normally
required to be refilled after a few years.
(e) Flame point should be above 55°C.
The physical properties of solvent vapour and air are given in
Table 13.1.
9 21 22 13
12
20
14 10
16
+–
4
1
18
2
11 5
3 19
19 6
2
17
7
8
15
1. Evaporator 12. Vacuum valve

2. Condenser 13. Aeration valve
3. Collecting tank 14. Filling valve for evaporator
4. Vacuum tank of transformer 15. Stop cock
5. Vacuum pumping group 16. Bypass valve
6. Vacuum pump 17. Reversing valve
7. Conveying pumps for solvent 18. Drainage pump
8. Storage tank 19. Filters
9. Solvent vapour return valve 20. Pressure switch-1
10. Solvent vapour inlet valve 21. Pressure switch-2
11. Valve for condensate 22. Pressure switch-3
Figure 13.5 Schematic of vapour drying system.
The following solvents meet the above characteristics and are

normally used in vapour phase drying systems.
Shellsol H (Shell)
Somenter T (Esso)
Varsol 60
Varsolene 60
Essovarsol 60 E
Torr 1000
5
4
3
2
H2O
100
5
4
3
2
Vapour-pressure
Solvent
10
5
4
3
2
1
20 40 60 80 100 120 140 160°C
Saturation temperature
Figure 13.6 Vapour pressure curve of solvent and water.
Table 13.1 Physical Properties of Solvent Vapour and Air
Sl. Physical properties Hydrocarbon Air

No. solvent
1. Specific density 0.785 g/cm3 1.25 kg/m3
(liquid) (gaseous)
2. Molecular weight 160 29
3
3. Heat of Vaporization 306 ¥ 10 W s/kg —
4. Specific heat 2.09 ¥ 103 W s/kg°C 1 ¥ 103 Ws/kg°C
(liquid) (gaseous)
5. Inlet temperature in 130°C 110°C
vacuum vessel
6. Outlet temperature from 90°C 90°C
vacuum vessel at start of
heating
7. Vapour pressure at 130°C 140 torr —
3
8. Energy provided per mole 62.7 ¥ 10 W s 581 ¥ 103 W s
3
9. Energy provided per mole 179 m at 130°C 31.4 m3 at 110°C
10. Energy released per m3 351 ¥ 103 W s 18.4 ¥ 103 W s
13.3.2 Requirements for Plant and Transformer
As the heat carrier vapour used in the vapour phase drying system is
inflammable, necessary precautions must be taken to avoid any
possibility of explosion. Maximum allowed leakage rate for a vacuum
vessel is 15 torr litre/s. Due to this reason, it is recommended that
insulation resistance and dielectric dissipation factor of windings be
measured during fine vacuum stage, only when the pressure is less
than 3 torr. Explosion is likely to materialize only in the event of a
sudden, massive leakage of atmospheric air and if temperature and
absolute pressure inside the vessel reach to the potential explosion
range. Figure 13.7 shows the potential explosion range for a mixture
of heat carrier vapour and air.
Torr 10 Atmospheric-pressure 760 torr

6
4 A
2
B
102
6
Absolute-pressure
4
Water
2
101
6
4
2
Solvent
100
6
4
20 30 40 50 60 70 80 90 100°C
Temperature
Figure 13.7 Explosion limits for a mixture of air and solvent.

(a) lower limit (b) upper limit.
Care should be taken to ensure that paints applied on inside and

outside surface of tank, end frame, clamp plate, copper bus-bar and
other ferrous components are compatible with solvent vapours at
130°C. Since, it is desired that insulation be impregnated just after
vacuum drying without exposing the winding to atmosphere for
tanking, suitable paints which have no reaction with solvent vapours
are to be applied on all parts, instead of conventional paint system.

Further, transformer is required to be designed such that outlet for
condensed solvent from all possible areas in the tank is available and
there is no possibility of large amount of solvent being trapped in
intricate insulation arrangement.
13.3.3 Drying Process
Transformer can be dried in a vacuum vessel or by evacuating its

tank, provided leakage rate is within required limits specified in Sec.
13.3.2. When the former is followed, core and windings assembly is
dried either alone or in its tank. In the latter case, an efficient heat
insulated compartment with hot air circulation system is required to
be provided to eliminate heat loss during drying and vacuum is
drawn in the transformer tank. Sequence of operation for VPD in a
vacuum vessel is described in the following paragraphs.
The drying process takes place in four stages indicated below, and
as shown in Fig. 13.8.
Phase I Phase II Phase III Phase IV

Setting-up Heating up Pressure Fine vacuum
reduction
1 ata
2
1
Pressure (torr)
Temperature
1. Temperature of transformer 3. Pressure (torr)

2. Temperature of heat conveyer 4. Amount of separated water
Figure 13.8 Stages of the vapour phase drying process.

(a) Preparation
(b) Heating up and drying
(c) Pressure reduction
(d) Fine vacuum
(a) Preparation (Setting up)

The entire evaporator and condenser system is first evacuated to an
absolute maximum pressure of 5 torr by the leakage air vacuum
pump, before drawing the solvent into the evaporator and heating
it to the required temperature of 130°C. The vacuum vessel valves
remain closed during this operation.
In parallel with the above preparatory steps, the vacuum system
evacuates the vacuum vessel containing the core and windings
assembly to approximately 5 torr. For draining the condensate
from the diverter switch oil compartment of on-load tapchanger,
wherever it is processed along with the transformer, the drain plug
in the bottom of the compartment is opened. Vessel floor is at a
descending slope of 1 : 100 towards the drainage system, so that no
condensed solvent remains inside the vessel. If tank containing core
and windings assembly is loaded into the vessel on a horizontal
trolley, tank is also kept at a slope of 1 : 100 to drain out the solvent
from the tank.
(b) Heating up and Drying

After the vessel is evacuated, vessel heating is started and this
heating is continued till the end of fine vacuum phase. The vessel
valves are opened at this stage, admitting vapours of heat-carrier
in the vessel, most of which condense on the cold surfaces of the
transformer. The condensed heat carrier is pumped back to the
evaporator through a filter. Heat released by condensation gradually
warms up the insulation and mass of component. Heat carrier
vapour pressure in the vessel and insulation moisture vapour
pressure both increase with rising temperature. As the water vapour
pressure is considerably higher than that of the heat carrier, insula-
tion moisture starts to evaporate at a comparatively low insulation
temperature. This produces mixtures of water vapour, leakage air
and heat carrier vapour inside the vessel, which is conveyed back to
the condenser via the vapour return in the condenser, whereas the
leakage air discharges to atmosphere through the vacuum pump.
The condensed water and heat carrier mixture is sent into the
collecting tank, in which its components settle out under gravity.

The water gets collected in the bottom of tank due to higher
specific weight, which is measured periodically and drained off. Final
insulation drying temperature of 120–125°C is maintained for the
time required, to ensure full moisture evaporation from the deeper
insulation layers. Longer the heating-up phase, the shorter is the
fine vacuum phase.
(c) Pressure Reduction

The vapour supply remains closed during this stage in which most of
the heat carrier absorbed by the insulation re-evaporate, condense
out in the condenser and is finally returned to the evaporator. This
phase is terminated when an absolute pressure of 15 to 20 torr is
reached in the vessel.
(d) Fine Vacuum

This is the final drying stage, which comes immediately after the
pressure reduction phase. It is same as conventional vacuum drying
(see Sec. 13.2.2) The vessel is evacuated by the main vacuum system
to a pressure not exceeding 0.1 torr. This phase is terminated
when water extraction rate is below the desired level and insulation
resistance and dissipation factor of windings become constant.
After drying of insulation by solvent vapours, other activities like
oil impregnation, soaking, draining, retanking redrying, completion
of fittings and oil circulation follow in the same manner as for
conventional drying (see Sec. 13.2) with the following exceptions:
(i) Since the transformer is at a temperature of 125°C at the end
of V.P.D., it is cooled to a temperature depending upon the
pressure in the vessel, such that oil is neither vapourized nor
oxidized during oil impregnation.
(ii) If only core and windings assembly is dried without tank, the
assembly is taken out from the vessel and is immediately
loaded into the tank. The tank is again kept in vessel and oil
is filled after evacuating vessel to the desired level and
removing any moisture absorbed by the insulation due to
exposure to the atmosphere.
13.3.4 Drying of Oil-impregnated Insulation

Vapour phase drying (V.P.D.) is ideal for drying transformers from
site after repairs, whose windings are exposed to atmosphere for a
longer period, since heat carrier washes the deteriorated insulating

oil out of the insulating materials. The high diffusion coefficient of
non-impregnated insulation is restored which is 20 to 30 times that
oil-impregnated insulation. Hence, total drying cycle is considerably
reduced and windings are cleaned of all sediment, fibres and other
impurities.
13.3.5 Paper De-polymerization
Detailed comparative measurements on paper specimens under

process conditions of V.P.D. and conventional drying have proved
that de-polymerization of paper is of the same order in these two
processing methods, as shown in Figs. 13.9–13.11. Short processing
times seem to be the major factor in reducing overall polymerization
in V.P.D. even though drying temperature is 130°C.
1400
A1
1300
1200
A2
1100
1000
0 24 48 72 96 120144 168192 h
H
Figure 13.9 Depolymerization of a paper for hot air drying at

atmospheric pressure.
A1 air temperature 90°C
A2 air temperature 110°C
H heating
13.3.6 Advantages of Vapour Phase Drying
Some of the advantages in the use of vapour phase drying of trans-

formers over the conventional system are as under:
(a) Uniform heating of the entire mass by penetration of the
hot solvent vapour ensures moisture extraction from the
innermost parts of the insulation, e.g. from precompressed
laminated pressboard clamping rings and crepe paper
insulation over HV leads, etc.
1400
A
1300
1200
0 24 48 72 96 120 144 168 192 h

H
V
Figure 13.10 Depolymerization of paper for hot air vacuum

drying vacuum tank. H heating V vacuum A air
temperature 95°C.
1400 B
1300
1200
0 24 48 72 h
H
V
Figure 13.11 Depolymerization of paper for vapour phase drying.
(b) Washing action resulting from the condensation of the heat

carrier vapour helps in removal of dust, dirt and fibres from
the innermost spaces in a transformer, and as such oil is not
contaminated and oil circulation time is considerably reduced.
(c) Ageing of insulation due to de-polymerization is comparable
or even less than that which occurs in the conventional drying.
(d) V.P.D. is more effective for drying transformers from site
after repairs.
(e) Drying time cycle before oil impregnation is reduced to the
extent of 30–40% of the usual system, thereby reducing the
transformer manufacturing cycle.
(f) Rapid build up of temperature without consequent damage to
insulation due to more uniform distribution of temperature in
various parts of the transformer, thereby reducing the heat
up time.
(g) Effective moisture extraction due to the deeper penetration of

hot solvent vapours and also higher water vapour pressure
differential (Refer Fig. 13.6).
REFERENCES
1. Mosser H.P. et al., Transformer Board Special Print of Scientia
Electrica, Translated into English by H. Weidmann, EHV.
Weidmann Limited, St. Johnsbury, Vermont, U.S.A., 1979.
2. Micafil News, Vapour–phase drying of EHV Transformer, MNV 46/le
November 1977, Micafil Limited 8048, Zurich, Switzerland.
3. D.P. Gupta et al., Vapour Phase Drying System for Transformers,
BHEL Journal, Vol. 3 No. 1, 1978.
4. IEC Publication, 422–1973, “Maintenance and Supervision Guide for
Insulating Oils in Service”.
C HAPTER 14
Testing of Transformers
and Reactors
P.C. Mahajan
M.L. Jain
R.K. Tiwari
Testing is an important activity in the manufacture of any

equipment. While certain preliminary tests carried out at different
stages of manufacture provide an effective tool which assures quality
and conformation to design calculations, the final tests on fully
assembled equipment guarantee the suitability of the equipment for
satisfactory performance in service. The basic testing requirements
and testing codes are set out in the national and international
standards. This chapter, however, is intended to cover the purpose
and the methodology of performing the tests.
With a view to cover detailed information about impulse and
partial discharge tests, which are of great importance, separate
sections have been devoted to these tests. A separate section has
been specifically devoted for short-circuit testing of transformers
which is a special test having significance in the transformer’s reli-
able service. Also, the specific requirements of reactor testing have
been dealt with in a separate section. All other tests and tempera-
ture-rise test on power transformer are described in the first section.
Thus, the chapter has been divided into following five sections:
Section I : Testing of power transformers
Section II : Impulse testing
Section III : Partial discharge testing
Section IV : Testing of reactors
Section V : Short-circuit testing of transformers.
SECTION I
14.1 Testing of Power Transformers
Preliminary tests are carried out on the transformer before it is put

into the tank. Final tests are carried out on completely assembled
transformer.
14.1.1 Preliminary Tests
Following tests are carried out in the works at different stages, be-
fore the core and coil assembly of the transformer is placed in its
tank. These checks help in detecting any fault at an early stage.
(a) Core Insulation

After the core is assembled, 2 kV test is done to ensure that the
insulation between clamp plates, core bolts and core is adequate.
(b) Core Loss Test

This is conducted on the core assembly to ensure its soundness.
Some turns are wound over the core and it is energized at normal
flux density. Coreloss and magnetizing current are noted and
compared with design value.
(c) Check of Ratio, Polarity, Vector Relationship and Winding

(c) Resistance of Transformer Assembly
Ratio test is conducted to ensure the correctness of voltage ratio
between different windings on each tapping. The tolerance allowed
for ratio is ±0.5% of the declared ratio or ±10% of the percentage
impedance voltage, whichever is smaller. The latter tolerance limit
is not applicable for auto-transformers and booster transformers,
where impedance value is small. In order to get accurate ratio, a
ratiometer is employed. It also indicates the polarity of transformer
windings.
For a three-phase transformer, it is more usual to carry out a
vector relationship test, in which one of the high-voltage and
low-voltage line terminals are jointed together as shown in Fig. 14.1.
Three-phase 400 V supply is connected across high voltage line
Testing of Transformers and Reactors 293
2U
2V
2W, 1U
3 phase
1N 40 V, 50 Hz
supply
1V
1W
Figure 14.1 Vector relationship test for star-delta (YN, d11)

cnnected step-down transformer.
terminals. Voltages between terminals 2U–1N, 2V–1N, 2W–1N, 2V–

1V, 2W–1W and 2V–1W are measured.
For YN, d11 vector relationship
2U–1N > 2V–1N > 1U–1N
and 2V–1W > 2V–1V or 2W–1W
The vector relationship for any other group can be checked in a
similar manner.
The dc resistance of each winding is measured by Kelvin’s double
bridge to check that there is no faulty joint.
(d) Preliminary Load Loss and Impedance Voltage Measurement

Preliminary load loss and impedance measurements at reduced
current are carried out to ensure that these are within guaranteed
limits.
14.1.2 Final Tests
The completely assembled transformer is tested in accordance with

the International Standards. The tests comprise the following:
(a) Routine tests
(i) Measurement of winding resistance
(ii) Measurement of voltage ratio and check of voltage vector
relationship
(iii) Measurement of impedance voltage (principal tapping) short
circuit impedance and load loss
(iv) Measurement of no load loss and current

(v) Measurement of insulation resistance
(vi) Dielectric tests
(vii) Tests on-load tapchangers (where appropriate)
(b) Type tests
All the tests listed as above and the temperature rise test and
Dielectric type tests.
(c) Special tests
(i) Dielectric special tests.
(ii) Measurement of zero-sequence impedance of three-phase
transformers
(iii) Short-circuit test
(iv) Measurement of acoustic noise level
(v) Measurement of harmonics of the no load current
(vi) Measurement of power taken by the fans and oil pumps
(vii) Measurement of capacitances between windings to earth
and between windings.
(viii) Measurement of transferred surge voltage on low voltage
windings.
(ix) Measurement of insulation resistance to earth of the
windings, or measurement of dissipation factor (tan delta)
of the insulation system capacitances.
(a) Routine Tests

(i) Measurement of winding resistance. For calculation of I2R-losses
in the winding, it is necessary to measure dc resistance of each
winding. The resistance measurement should be done after the
direct current circulating in the winding has reached a steady state.
In some cases this may take several minutes depending upon the
winding inductance.
Temperature of the winding must be stable and for this reason,
this test is carried out usually before load loss measurement. The
average oil temperature is determined as the mean of top and bot-
tom oil temperatures and is taken as average winding temperature.
(ii) Measurement of voltage ratio and check of voltage vector group.
These tests are conducted as described in para 14.1.1 (c) to check
that all connections to the bushings, tapchangers, etc., have been
made correctly during final assembly.
(iii) Measurement of impedance voltage and load loss. The load loss
comprises the sum of the I2R-losses in the windings and the stray
losses due to eddy currents in conductors, clamps and tank. The stray
losses vary with frequency. It is thus important to give supply to the
transformer at the rated frequency. The load loss and impedance
voltage are guaranteed at 75°C but are measured at the ambient
temperature of the test room. Measured load loss is corrected
to reference temperature (75°C). It is known that I2R-losses are pro-
portional directly to resistance, which varies with the temperature.
The stray losses vary inversely with the temperature.
The test is carried out by short circuiting , usually the LV winding
and by supplying the impedance voltage to HV winding. The
measured power will also include small core-loss. Since the supply
voltage during the test is a small fraction of normal voltage, this loss
can be ignored. However, for a high impedance transformer (i.e. Z >
15%) The core-loss may become appreciable and can be deducted
after separately measuring it at impedance voltage.
As per IS : 2026 (Part I)—1977, the measurements can be made at
any current between 25 to 100%, but preferably not less than 50% of
the rated current. Load loss and impedance voltage can be corrected
for rated MVA as below:
Computed loss at rated current
2
= Measured loss at test current ¥
FG Rated current IJ
H Test current K
Test voltage Rated current
%Z = ¥ ¥ 100
Rated voltage Test current
While measuring load loss and impedance at different tap positions,
readings should be taken quickly, and the interval between the mea-
surements at different taps should be adequate to avoid significant
errors due to momentary temperature rise of the windings. The
difference in temperature between the top oil and bottom oil should
preferably be small enough to enable the average temperature to be
determined accurately.
Three-wattmeter method should be used instead of two wattmeter
method to avoid large value of wattmeter multiplier constant.
The power factor during load loss test can be less than 0.1 and
wattmeters suitable for such low power factor should be used.
Following tolerances (Table 14.1 are applicable as per IS : 2026,

for specified values of losses and impedance voltage.
(iv) Measurements of no-load loss and current. The measurement of
no-load loss and current is important not only for the purpose of
assessing the efficiency of the transformer, but also as a check that
the high-voltage tests have not caused any damage to winding insula-
tion. For large transformers, therefore, no-load loss measurement is
carried out before and after completion of the dielectric tests.
Table 14.1 Tolerances
S. No. Item Tolerance

(a) Total losses +10% of the total losses
(ii) Component losses +15% of each component loss,
provided that the tolerance
for total losses is not exceeded
(b) Impedance voltage at rated
current
(i) Principal tapping
(1) two-winding transformers ±10% of the declared impedance
voltage
(2) Multi-winding transformers ±10% of the declared impedance
voltage for one specified pair of
windings
±15% of the declared impedance
for a second specified pair of
windings
(ii) For tappings other than the Tolerance shall be increased by
principal tapping a percentage equal to half the
difference in tapping factor (per-
centage) between the principal
tapping and the actual tapping
Measurement of no-load loss is carried out at rated frequency

feeding usually LV winding. Since the no-load current is very small,
the I2R-losses in the windings will be negligible. The power factor,
while measuring no-load loss for medium power transformer, is
generally around 0.3. The power factor for large units, especially
when working at higher flux densities is about 0.1.
The core-loss consists of hysteresis and eddy current losses.
The hysteresis loss is dependent on average value used and the eddy
current loss on rms value of supply voltage. Two voltmeters are used
during the test, a bridge rectifier type to indicate average voltage,
and a dynamometer type to indicate rms voltage. The supply
voltage is set so that the specified value is indicated on the average
voltmeter. With this, hysteresis component of no-load loss will be
measured correctly, whilst the eddy current loss will be either lower
or higher than the true value, depending upon the form factor of the
supply voltage. Since the ratio between the two components is
known for any particular quality of core steel, losses are corrected by
using the following formula:
Pm
P =
P1 + KP2
where P = no-load loss for sinusoidal voltage
Pm = the measured no-load loss
P1 = ratio of hysteresis losses to total iron loss
P2 = ratio of eddy current loss to total iron loss, and
2
K =
FG rms voltage IJ
H 1.11 ¥ average voltage K
For normal flux densities and frequencies of 50 Hz and 60 Hz, P1
and P2 are each 0.5 for grain-oriented steel, and 0.7 and 0.3 respec-
tively, for non-grain-oriented steel.
(v) Measurement of insulation resistance. Insulation resistance is
measured between all windings and the tank with a megger. The
insulation resistance varies inversely with the temperature.
Thus the oil temperature is also recorded. Sometimes insulation
resistance values at 15th and 60th second are noted to determine
polarization index of the insulation system.
(vi) Dielectric tests.
(1) Separate source voltage withstand test. This test is intended
to check the adequacy of main insulation to earth and
between windings.
The line terminals of the windings under test are
connected together and the appropriate test voltage is applied
to them while the other windings and tank are connected
together to the earth. The value of test voltage for fully-
insulated winding is indicated in Table 14.2. Windings with
graded insulation, which have neutral intended for direct
earthing, are tested at 38 kV.
The supply voltage should be nearly sinusoidal. The peak

value of the voltage is measured and for this a digital peak
voltmeter associated with capacitive voltage divider is
employed. The peak value divided by 2 shall be equal to the
test value. The duration of application of test voltage is 60 s.
(2) Induced over-voltage withstand test. The test is intended to
check the inter-turn and line end insulation as well as main
insulation to earth and between windings.
For transformers with uniformly insulated windings, the
test voltage is twice the corresponding rated voltage.
For transformers with non-uniformly insulated HV
windings, which are usually designed for systems with
highest voltage Um ≥ 72.5 kV, the values of test voltage are
indicated in Table 14.2 (below the dotted line).
For test voltage up to 66 kV, three-phase transformers are
generally supplied direct from a three-phase source.
For higher values, it is the usual practice to raise each HV
terminal in turn, to specified test voltage by applying single-
phase voltage to LV winding. The neutral terminal may be
raised to any appropriate voltage, so as voltage per turn of
the phase under test shall be at least twice the normal value.
In order to avoid core saturation at the test voltage, it is nec-
essary to use a supply frequency higher than the normal. Fig-
ure 14.2 indicates a typical induced over-voltage test circuit
for a generator transformer.
When the frequency is chosen in the range of 150 to 240
Hz, the capacitive reactance of transformer is appreciably
reduced and it draws significant capacitive current at test
voltage, which causes heavy loading on the generator set.
The loading on the generator can be reduced by connecting a
variable reactor across the generator terminals.
Self-excitation of generator and series resonance between
the capacitive load and any series inductance in the test
circuit must be avoided by proper selection of test frequency.
As a precaution, a sphere gap can be connected across
the high voltage winding, so that it will spark-over if any
excessive voltage appears on the terminal.
Test duration is determined by the following formula:
120 ¥ Rated frequency
Test duration in seconds =
Test frequency
Table 14.2 Rated Withstand Voltages for Transformer Windings with

Highest Voltage for Equipment Um up to and Including
420 kV
Highest voltage for Rated short-duration power frequency

equipment kV (rms) withstand voltage kV (rms)
1.1 3
3.6 10
7.2 20
12 28
17.5 38
24 50
36 70
52 95
72.5 140
123 185 230
145 230 275
245 360 395
420 570 630
1N 1U 1V 1W
395 kV
3 3 3
4 5 4 5 4 5
6 7 6 7 6 7
8 9 8 9 8 9
10 10 10
35.67 kV
2U 2V 2W 150 Hz
2U2 supply
2.43
2.43 kV kV
–4.86 kV
Figure 14.2 Test circuit for induced over-voltage test (the arrows
indicate the direction of voltage).
The high voltage is normally measured by means of a

digital peak voltmeter associated with capacitive voltage
divider.
(b) Type Tests
(i) Temperature rise test. The test is conducted to confirm that under
normal conditions, the temperature rise of the windings and the
oil will not exceed the specified limit. Temperature rise limits for
oil-immersed type transformers are indicated in Table 14.3.
Table 14.3 Temperature Rise Limits for Oil Immersed Type Transformers
S.No. Part Temperature rise (°C) Condition

External cooling medium
Air Water
1. Windings 55 60 When the oil circulation is
natural or forced non-directed
60 65 When the oil circulation is
forced directed
2. Top oil 50 55 When the transformer is sealed
or equipped with a conservator
45 50 When transformer is neither
equipped with a conservator
nor sealed
The temperature rises are measured above the temperature

of the cooling air for all types of transformers except those water
cooled. In the latter case, the temperature rise is measured above
the inlet water temperature.
(ii) Temperature rise for top oil. Normally, LV windings of the
transformer under test is short circuited and a voltage of such a
value is applied to HV winding that power input is equal to no-load
loss plus load loss corrected to a reference temperature of 75°C. For
multi-winding transformers, the temperature of the top oil refers to
the specified loading combination for which the total losses are
the highest. The total losses are measured by the three-wattmeter
method in the circuit and maintained constant until the top oil rise
has reached a steady value.
If the total losses cannot be supplied due to plant limitations,

losses not less than 80% of the total loses are supplied and following
correction factor is applied to top oil temperature rise
x
FG Total losses IJ
H Test losses K
The value of x for
natural air circulation : 0.8
Forced air circulation and water cooling : 1.0
During the temperature rise test on an oil-immersed type
transformer, hourly readings of the top oil temperature are taken by
means of a thermometer placed in a pocket in the transformer top
cover. The temperature of oil at inlet and outlet of the cooler bank is
also taken hourly and mean oil temperature is determined. Ambient
temperature is measured by means of thermometers placed around
the transformer at three to four points situated at a distance of 1 to 2
m from and half-way up the cooling surface of the transformer.
The thermometers are inserted in oil cups, which are filled with
transformer oil.
(iii) Duration of temperature rise test The test is continued until the
requirement of one of the following methods have been met:
Method (a)
The test should not be regarded as complete until the temperature
rise increment of top oil is less than 3°C in 1 hour. The method
shown in Fig. 14.3 shall be employed for the determination of the
final oil temperature rise.
Method (b)
It should be demonstrated that the top oil temperature rise does
not vary more than 1°C per hour during four consecutive hourly
readings. The last reading is taken for determination of top oil rise.
(iv) Winding temperature rise. When top oil temperature rise is
established, the current is reduced to its rated value and is main-
tained for 1 hour to allow the winding to attain normal temperature.
If the rated current cannot be supplied, the tests can be performed
with a current not less than 90% of the rated current. Following
correction factor is applied to determine the winding gradient
corresponding to rated current.
Temperature
rise (°C)
Final top oil rise
Dq3
Dq3 Dq2
Dq2 Dq1
Dq1
Temperature
Increment 0 Dt Dt Dt Dt Dt
Time in Hr.
Figure 14.3 Method for determining the final temperature rise of oil.
y
Winding gradient = Test gradient ¥
FG Rated current IJ
H Test current K
The value of y for
natural and forced non-directed oil circulation : 1.6
forced-directed oil circulation : 2.0
At the end of the test, the supply is switched off. The cooling fans
or water pumps should be stopped but the oil pumps should remain
running. The short-circuit connection is removed. The value of hot
resistance of the winding is measured by Kelvin’s double bridge or by
Tettex resistance measuring equipment. A certain time, about 3 to 4
minutes, usually elapses between switching off the power supply and
taking the first reading, during which the resistance of winding will
be decreasing. In order to determine the temperature of winding at
the instant of power switch-off, the resistances are measured at in-
tervals over a period of about 15 min. Graph of hot resistance versus
time is plotted, from which winding resistance (R2) at the instant of
shut-down can be extrapolated in the manner shown in Fig. 14.4.
From this value, q 2, the winding temperature at the instant of
shut-down can be determined as below:
R2
q2 = (235 + t1) – 235
R1
R2
Hot resistance
q
DR1
DR1 DR2
DR2 DR3
DR3
Dt Dt Dt Dt
DR
Resistance O Time, minutes
increment
Figure 14.4 Method for determining the winding resistance at

the instant of switching off the supply.
where, R1 is the cold resistance of winding at temperature t1, 235

being reciprocal of the temperature coefficient for copper. Winding
temperature rise can be determined by subtracting mean ambient
temperature from q2 and adding the drop in mean oil temperature
rise, if any, from steady-state condition (that is, before current is
reduced to rated value) to shut-down condition.
(v) Lightning impulse test. It is dealt with in Sec. II.
(vi) Switching impulse test. It is dealt with in Sec. II.
(vii) Partial discharge test. It is dealt with in Sec. III.
(c) Special Tests

(i) Measurement of zero-sequence impedance of three-phase trans-
former. The zero-sequence impedance is measured on star-connected
windings, which have a earthed neutral, in order to determine the
current which will flow in the event of a line-to-earth fault. Reluctance
path for zero sequence flux is different in three-phase three-limb
core and three-phase five-limb core and hence the value of zero se-
quence impedance depends upon type of core used in a transformer.
Usually the value of zero sequence impedance lies between 80–90%
of positive sequence impedance for transformer having three-limb
core, whereas it is between 90–100% of positive sequence impedance
for transformer having five-limb core construction.
Figure 14.5 shows the connection for carrying out this test on a
delta-star connected transformer. The line terminals on the star
connected windings are joined together and single-phase supply is
applied between these and the neutral point, the delta terminals
being left floating during this test.
Supply 1N 1U 1V 1W
2U 2V 2W
Figure 14.5 Test circuit for the measurement of zero sequence

impedance on a star-delta transformer.
For transformers with more than one star-connected windings

and neutral terminal, additional measurements of zero-sequence
impedances are made, in which the line terminals and the neutral
terminal of the other star-connected winding are connected together.
Auto-transformers with a neutral terminal are treated as normal
transformers with two star connected windings.
Zero-sequence impedance in ohms can be obtained as below
Z0 = 3 V/I where V is the test voltage and I is the test current.
(ii) Short-circuit test. It is dealt with in Section V.
(iii) Measurement of the harmonics of no-load current. The harmon-
ics of the no-load current in all the phases are measured by means
of harmonic analyzer at rated voltage and the magnitude of the har-
monics is expressed as a percentage of the fundamental component.
(iv) Measurement of acoustic sound level. Acoustic sound test mea-
sures the average sound level generated by the transformer, when
energized at rated voltage and frequency at no-load with cooling
fans and pumps in operation. The main sound comes from the core
vibration, since there are no moving parts in the transformer. Noise
level measurements are taken in accordance with NEMA-TR1. The
measurement details are covered in Sec. IV
SECTION II
14.2 Impulse Testing
Lightning is probably the most common cause of flashover on over-

head transmission lines. Two mechanisms can be distinguished. In
the first, the lightning stroke makes a direct contact with a phase
conductor producing a voltage on the line in excess of the impulse
voltage level and in the other, the stroke makes contact with an
earth wire or tower and the combination of tower current and tower
impedance produces a voltage near the tower top sufficient of pre-
cipitate a so called back flashover. The terminal equipments of high-
voltage transmission lines experience lightning impulses in service.
Insulation is one of the most important constituents of a
transformer. Any weakness of insulation may result in failure of a
transformer. A measure of the effectiveness with which insulation
performs is the dielectric strength. The power frequency tests alone
are not adequate to demonstrate the dielectric strength of trans-
formers. The distribution of impulse voltage stress along the trans-
former winding is different from the power-frequency voltage distri-
bution. The power-frequency voltage distributes itself throughout
the winding on a uniform volts per turn basis. Impulse voltage is
initially distributed on the basis of the winding capacitance and
finally the inductance. Thus, it is necessary to ensure adequate
dielectric strength of transformers for impulse conditions.
Switching impulses occur during all kinds of switching operations
in the system, for example by switching a transformer ‘on’ or ‘off’ a
system, switching of distribution line, etc. The magnitude and form
of impulses produced differ from case to case. The magnitudes of
switching impulses occurring in the network are proportional to the
network voltage. The maximum voltage can be about 3.5 times the
service voltage (the ratio of the maximum switching impulse voltage
to the maximum peak phase-to-earth service voltage). This ratio is
as low as 2.1 on some EHV systems. The need for switching impulse
tests is based on the possibility that switching impulses in service
can cause insulation damage to a transformer designed for greatly
reduced insulation level. The switching impulse tests are worth-
while for transformers with reduced insulation level. Studies and
researches on insulation arrangements have shown that, a trans-

former designed without taking switching impulses into account,
can withstand switching impulse voltages equivalent to 83% of the
corresponding basic impulse level (BIL).
14.2.1 Lightning Impulse
Impulse tests are made with waveshapes which simulate conditions

that are encountered in service. From the data compiled about natu-
ral lightning, it has been concluded that system disturbances from
lightning can be represented by three basic wave shapes—full waves,
chopped waves and front of waves. It is also recognized that lightning
disturbances do not always have these basic wave shapes. However,
by defining the amplitude and shape of these waves, it is possible to
establish a minimum impulse dielectric strength that a transformer
should meet.
If a lightning disturbance travels some distance along the line
before it reaches a transformer, its waveshape approaches that of
the full wave (Fig. 14.6). It is generally referred to as 1.2/50 wave. A
wave travelling along the line might cause flashover across an
insulator after the peak of the wave has been reached. This wave is
simulated by a chopped wave which is of the same magnitude as the
full wave, as shown in Fig. 14.6. If a lightning stroke hits directly at
or very near to a transformer terminal, the impulse voltage may rise
steeply until it is relieved by a flashover, causing sudden, very steep
collapse in voltage. This condition is represented by the front of wave
(Fig. 14.6).
(c) Front of wave

(b) Chopped wave
1.0 (a) Full wave

Voltage
0.0
Time
Figure 14.6 Lightning impulse wave shapes.

These three waves are quite different in duration and in rate of

rise and decay of voltage. They produce different reactions within
the transformer winding. The full wave, because of its relatively long
duration, causes major oscillations to develop in the winding and
consequently produce stresses not only in the turn-to turn and sec-
tion-to-section insulation throughout the winding, but also develop
relatively high voltages compared to power frequency stresses
across the large portions of the winding and between the windings
and the ground.
The chopped wave, because of its shorter duration, does not allow
major oscillations to develop fully and generally does not produce as
high voltages across the large portions of the windings. However,
because of the rapid change of voltage following flashover, it pro-
duces higher turn-to-turn and section-to-section stresses.
The front of wave is still shorter in duration and produces still
lower winding-to-ground voltages within the winding. The rapid
change of voltage on the front followed by a flashover produces high
turn-to-turn and section-to-section voltages very near the line-end of
the winding. The front-of-wave test is not carried out on transform-
ers but is a very important test for lightning arresters.
14.2.2 Switching Impulse
Studies on network models and practical measurements in the net-

work regarding the form and magnitude of unipolar voltages due to
switching operations reveal that these over-voltages have front
times of several hundred microseconds. The form of over-voltage
may be a periodically damped or oscillating one.
The IEC-60060 have adopted, for switching impulse test, a long
wave having front time 250 ms and time to half value 2500 ms with
tolerances. The American Standard recommends the duration of
voltage wave at 90% of the crest value of at least 200 ms. The actual
time to crest shall be greater than 100 ms and the time to first zero-
voltage on the wave tail shall not be less than 1000 ms, except where
core saturation causes the tail to become shorter. In view of the
difficulties in generating these wave-shapes in laboratories and more
to minimize the core saturation during testing of transformers, IEC-
60076 Pt. III has specified a switching impulse wave having a virtual
front time of at least 20 ms, duration above 90% of the specified
amplitude of at least 200 ms and a total duration to the first zero

passage of at least 500 ms. The time to crest has been chosen so that
the voltage distribution within the winding will be essentially uniform.
The dielectric strength of air, particularly for a positive unipolar
impulse is dependent on the front time of these unipolar impulses,
and feature a pronounced minimum in the range of 100–300 ms.
Therefore, to avoid flashover on the external insulation of the trans-
former, it is preferred to conduct the test with negative polarity im-
pulse.
The switching impulse voltages are of long duration and generally
do not cause any non-uniform voltage distribution along the winding,
but they affect the insulation to ground, between windings and
phases and the external insulation of the bushings. During the
switching impulse test, the open circuited low-voltage winding on
the same core limb, as the impulsed HV winding, is automatically
subjected to impulse voltage stresses. These voltages are induced
proportional to the transformation ratio and the flux distribution.
Further tests on low-voltage windings are thus not essential.
14.2.3 R.S.O. Techniques
It is not possible to measure the voltage in different parts of a coil

during the transformer impulse test. The behaviour of windings can
be analytically calculated with the use of a computer; alternatively,
the voltage distribution in windings can be obtained with recurrent
surge oscillograph techniques. The recurrent surge generator is a
low-voltage equivalent of a high-voltage impulse generator with
lumped parameters. Repetitive pulses of low voltage are applied to
the winding terminal from a recurrent surge generator. The volt-
ages induced and their shape in different parts of the winding and
other windings can be measured and recorded with an oscilloscope.
The recurring impulse wave gives a completely still picture on the
oscillograph screen. The applied waveform to give desired front and
tail duration can be adjusted. It is thus possible to measure the volt-
age distribution in the winding and the voltage differential between
the windings, thus allowing a suitable choice of the insulation. The
R.S.O. test, being a low-voltage test, allows the study of behaviour of
windings exposed to atmospheric surges. The R.S.O. test is also done
on completely assembled transformer to find out circuit parameters
for generation of desired impulse waveshapes.
14.2.4 Generation of Impulses and Test Circuit
Lightning and switching impulse test is conducted by means of an

impulse generator by direct voltage application on the winding under
test.
Impulse waves are generated by an arrangement that charges a
group of capacitors in parallel and then discharges them in series.
The magnitude of the voltage is determined by initial charging volt-
age, the number of capacitors in series at discharge and the regula-
tion of the circuit. The waveshape is determined largely by the con-
stants of the generator and the impedance of the load. The impulse
generators built according to well known Marx multiplier circuit are
capable of generating different unipolar over-voltages from several
microseconds to a few hundred microseconds. For inter-phase test-
ing of EHV transmission towers and external insulations of EHV
substations with switching impulses of very long wave front from
1000 to 5000 ms, the conventional impulse generation technique be-
comes less effective due to problems in triggering all generator
stages, arc extinction in generator spark gaps, etc. These waves can
also be generated using cascade connected testing transformers by
either discharging a capacitor bank through the primary winding of
the transformer or by energizing the transformer during a short
time period, typically less than one-half cycle of the power frequency.
Principles of Waveshape Control

The impulse waveshape is influenced by the generator capacitance
(Cg), internal and external series resistances (Rsi and Rse), parallel
resistance (Rp), loaded capacitance (Ci) (including capacitance of the
voltage divider, the chopping equipment, if used) and the effective
capacitance and inductance of test object (Ct, Li). The simplified
circuits for the control of the wave front and wave tail durations for
impulse test are given in Fig. 14.7.
The front time will be approximately
C g ◊ Ci
T1 ~
- 3 ◊ Rs ◊ (1)
C g + Ci
For Cg >> Ci,

T1 ~
- 3 ◊ Rs ◊ Ci (2)
Rs Rsi
Cg Ci Cg Ci
Rp
(a) Wave front control (b) Wave tail control
Figure 14.7 Wave shape control circuits.
The time to half value will be approximately

T2 ~
- 0.7 (Rs + Rp) (Cg + Ci) (3)
for Rs << Rp and Cg >> Ci,
T2 ~
- 0.7 Rp ◊ Cg (4)
For purely capacitive loads, the front and tail durations can be
adjusted according to the above equations. However, in the case of
transformer, the effective transformer capacitance Ct included in the
total load capacitance is a different physical quantity for front and
tail considerations. For the front time, Ct is defined as Ct = Cs ◊ Cg ¢ ,
where Cs is the winding series capacitance and Cg¢ is the winding
earth capacitance. For the wave tail, Ct is defined as Ct ~
- Cg¢.
14.2.5 Lightning Impulse Test Circuit
For testing of transformers, the basic arrangement of test circuit has

been shown in Fig. 14.8. The physical arrangement can be sub-di-
vided into three major circuits:
— The min circuit, which includes the impulse generator, wave-
shaping components and the test object,
— The voltage measuring circuit, and
— The chopping circuit, where applicable.
The standard lightning impulse is a full lightning impulse having a
virtual front time of 1.2 ms and a virtual time to half-value of 50 ms
with tolerances of ±3% on peak value, ±30% on front time and ±20%
on time to half-value. While testing high voltage windings of large
transformers, viz. 400 kV, 250 MVA, it becomes difficult to obtain the
Rsi R
se
ng
Impulse Ma uri
in e as
generator Ci m
rcu ge
Rp it
o lta cuit Z Voltage
Cg Ci V cir 1 divider
g
pin
h op Test U
C it
cu object
cir Ct, Lt Z2
Sh
Chopping
Reference earth
gap
Cg Generator capacitance
Ct Load capacitance (can also be used as divider)
Ct Effective test objest capacitance
Lt Effective test objest inductance
Rsi Internal series resistance
Rse External series resistance
Rp Parallel resistance
Sh Shunt for current measurements
Figure 14.8 Impulse test circuit.
wave front within the specified limits. This is due to the large
capacitance of the transformer, the inherent self-inductance of the
generator and the leads connecting the generator and transformer.
On low-voltage winding of these transformers, it becomes difficult to
obtain the specified wave tail due to extremely low impedance of
these windings.
The effective impedance of transformer controls the wave tail du-
ration. The effective impedance can be varied by different terminal
impedances at the non-impulsed winding terminals. If they are iso-
lated, the effective impedance will be maximum. However, they can-
not be left isolated because high voltages are induced in these wind-
ings. It is desirable that the voltage on terminals that are not being
tested is limited to 75% of its BIL. To ensure this, non-impulsed

terminals of the windings are earthed directly of through resistors
(maximum up to 400 W).
The full-wave test sequence consists of application of a reduced
full-wave and three full waves. The dielectric stress distribution in a
transformer under impulse test depends on the tapping connection
and the design in general. Thus, for lightning impulse test, different
tappings are selected for the tests on the three phases of a three-phase
transformer or the three single-phase transformers of a three-phase
group, for example, the two extreme tappings and the principle
tapping.
14.2.6 Test with Impulse Chopped on the Tail
Stresses set up in windings by chopped waves are quite different

from those set up by full waves. The chopped wave produce greater
stresses to the line portion of the winding, when compared to a full
wave due to the fast rate of collapse of voltage during chopping.
Different times to chopping will result in varying stresses in dif-
ferent parts of the winding, depending on the winding construction
and arrangement. It is thus not possible to state a time to chopping,
which is most onerous either in general, or for any particular
transformer, or reactor. The time to chopping between 2 and 6 ms is
considered to provide severe line-end stresses in the transformer
winding and has been specified for withstand test.
When measuring chopped impulses, the measuring system should
be accurate to record rate and duration of the voltage collapse and
the amplitude of reversed polarity. During the chopping, both the
rate of collapse and the amplitude of opposite polarity are largely
dependent on the geometrical arrangement of the chopping circuit
and on the impedance of the circuit. The IEC 60076–3 and other stan-
dards specify the amount of overswing to opposite polarity to be
not more than 30% of the amplitude of the chopped impulse. The
triggered type of chopping gap provides better consistency of the
time to chopping. The detection of failure during chopped wave tests
is more difficult than with full waves. Chopped wave impulses do not
give identical oscillations after chopping, unless the time to chop are
identical. The deviations on the wave shape before the chopping give
more positive indication of abnormality. However, the failure during
the chopped wave test are reflected on application of full waves,

immediately after the chopped wave and these current and voltage
oscillograms are compared with full wave oscillograms before
chopped wave application.
The standard sequence of application is:
— one reduced full impulse,
— one 100% full impulse,
— one reduced chopped impulse,
— two 100% chopped impulses,
— two 100% full impulses.
14.2.7 Switching Impulse Test of Transformer

The test circuit of direct application of switching impulses on the HV
winding is shown in Fig. 14.9. Unlike in the case of lightning impulse
test where the windings not being tested are grounded, for switching
impulse test these windings are left open circuited as the loading
impedance at short-circuit is low and prevents the application of
switching impulse voltages of the required duration and magnitude.
The iron core becomes magnetized due to long duration switching
impulses. The application of successive impulses causes a progressive
increase in the residual flux in the core. Sounds, which, under nor-
mal impulse tests would be related to defects in transformers may
occur during switching impulse tests due to magnetostriction.
The front of the wave depends mainly on the value of series resis-
tance Rse, the effective capacitance to ground of tested winding and
the divider capacitance. As long as the core is unsaturated, the
current flowing through the winding remains relatively low. After
the peak of the wave is reached, the electric charge is practically
conducted only through the parallel resistance Rp. The test voltage
remains practically constant. Thus,
df
U = (14.1)
dt
where, U = test voltage
f = magnetic flux
t = time
for f < fs (saturating flux)
f =U◊t (14.2)
where, U = peak value of the test voltage.
Rse 1U 1V 1W
Ct
Rp
Z
Cg
Sh
M
Voltage
Reference earth 2U 2V 2W
Current
Cg = Generator capacitance Sh = Shunt for current measurement

Rp = Parallel resistance 1U, 1V, 1W = High voltage terminals
Rse = Series resistance N = Transformer neutral
Ct = Capacitive voltage divider 2U, 2V, 2W = Low voltage terminals
Z = Surge impedance M = Iron core
The arrows indicate both direction and magnitude (one arrow is equivalent
to half the value of the magnetic flux and the respective voltage)
Figure 14.9 Circuit for switching impulse test on a transformer.
The flux f increases linearly and reaches after a certain time ts,
the value fs, the saturation flux for the corresponding test voltage.
The impedance of the winding and consequently the test voltage
collapses at saturation. The current in the winding increases rapidly
and reaches the peak value when voltage reaches zero. The higher
the test voltage, the faster is the saturation point reached and conse-
quently shorter becomes the wave duration. Core saturation is,
therefore, the limiting factor for the wave duration. Even a large
source of energy (i.e. high energy rated impulse generator) cannot
prolong the duration of the wave. Biasing magnetically the core
in the opposite direction helps in increasing the wave duration
considerably.
From Eq. (14.1)

f2
z U dt = z
f1
df = f2 – f1
Taking f2 as saturation flux and f1 as residual flux, the difference

is the factor which determines the area enclosed in the total wave
between the voltage trace and the time axis. For any particular test
voltage, the wave duration is proportional to the difference between
the saturation flux and residual flux. Thus, a residual flux of opposite
polarity can increase the wave duration of that particular test
voltage. This is accomplished by application of opposite polarity
switching impulse of reduced amplitude (Say 50 to 60 %).
One reduced full-wave (75%) and two full-waves of rated voltage
are applied on the winding under test. For fault detection, the re-
cording of voltage wave is done, neutral current can also be recorded.
Exact comparison of oscillograms is not possible as both the voltage
and current waves change in magnitude and duration.
14.2.8 Measurement and Recording of Impulses
To measure the amplitude and shape of the applied impulses which

have values ranging from a few tens to over thousands of kV and
duration 0.2 to 250 ms for the peak, special measuring equipments
are used. Oscillographs with high writing speeds and good accuracy
and voltage dividers with response time suitable for extremely fast
transients are required. Digital readout impulse peak voltmeters
are used for amplitude measurements. There are three basic types
of dividers that are suitable for impulse testing. The resistance
divider utilizes the principle that the voltages across a resistor
varies directly as the resistance, in the capacitance divider the
voltage varies inversely as the capacitance. The compensated
dividers are the combination of the two. The capacitance and com-
pensated dividers also serve as the load capacitor while generating
impulses.
For switching impulse test, capacitive types of voltage dividers
are preferred. Resistive voltage dividers have an influence on the
efficiency of the circuit and waveshape. They may also be thermally
overloaded.
The chopping of impulse wave can be obtained with a rod gap,

triggered sphere gap or with a multiple chopping gap. The rod gap
and triggered sphere gap lack consistency in chopping duration,
which can be obtained with a multiple chopping gap.
Oscillographic Recording
(i) Lightning impulse test. The applied voltage wave and one other
parameter, whose choice depends on the selection of method of
failure detection, are recorded. For best comparison, oscillograms
taken at reduced and full test levels should be recorded to give equal
amplitude by the use of attenuators at the oscilloscope.
Recording of Voltage
(a) Wave shape recording. The preferred sweep time for the wave
front record is 5–10 ms (longer sweep time may be required
when testing transformer neutrals) and for wave tail 50–100
ms.
(b) Test wave recording. For full waves, the sweep time should
not be less than 50 ms and the chopped waves should be
recorded at 10–25 ms sweep.
Recording of Current
The impulse current are normally the most sensitive parameters in
failure detection and record of current waves are the main criteria
of the test result. The use of more than one recording channel at
different sweeps gives better resolution. The criteria for selection of
sweep is
(a) to obtain a clear representation of the oscillations including
the high frequency components near the front of wave;
(b) to be able to detect any discrepancies occurring late in time.
Recording of current at 10 ms and 100 ms sweep covers the above
requirements in general.
(ii) Switching impulse test. During switching impulse test, only
the recording of applied voltage is required. The voltage record
will indicate any fault developed on winding under test or other
non-tested windings.
Recording of Voltage
(a) Waveshape recording. For the wave front record, a sweep
time 100–200 ms is used. For the wave tail record, by which
the time above 90% is determined, a sweep time of 1000–2000
ms is adequate.
(b) Test wave recording. The sweep time for test wave recording
should be long enough to encompass the first zero passage,
generally a sweep time of 1000–3000 ms is used. Logarithmic
time sweep is also used.
Recording of Current
A switching impulse current comprises three parts:
— an initial current pulse
— a low and gradually rising value of current coincident with
the tail of applied voltage
— a peak of current coincident with any saturation
It is usual to employ the same sweep time as used for voltage record.
14.2.9 Fault Detection
The detection of faults is the most important phase of impulse test-

ing. The detection of failure with CRO is most effective and sensitive.
It is based on the fact that an insulation failure will change the
impedance of the transformer, causing a variation in the impulse
current and in the voltage. Both the voltage and current oscillo-
grams are taken during impulse test. Reduced full-wave oscillogram
represents the characteristic waveshape of the transformer winding
and circuit at a voltage stress, when transformer is considered to be
healthy. Subsequent full-wave oscillograms are compared with
reduced wave oscillograms taken as reference. Full-wave oscillograms
prior to and after the chopped wave application match when winding
is sound.
For failure detection during lightning impulse test, three different
currents can be measured and used separately, or in combination as
shown in Fig. 14.10.
(a) The neutral current
(b) The capacitively transferred current
(c) The tank current
(c) (a) (b) (d)
(a) Neutral current (c) Tank current

(b) Capacitively transferred current (d) Transferred voltage
Figure 14.10 Methods of failure detection.
In addition, the transferred voltages to non-tested windings can

also be used for detection of failure.
For failure detection in switching impulse test, normally the mea-
surement of applied voltage is adequate, but the current flowing
through the winding to earth can also be used. For recording of cur-
rent, a non-inductive resistor is used as the impedance for producing
the voltage drop.
14.2.10 Interpretation Oscillograms

(a) Lightning Impulse
Interpretation of oscillograms is a skilled task and it is often difficult
to decide the significance of discrepancies, because of the large num-
ber of possible disturbance sources, while investigating discrepan-
cies. First, it should be checked that the test circuit and earthing are
not causing disturbances. Sometimes the recording oscillograph gen-
erate small disturbances which can be mistaken as a fault inside the
transformer. Secondly, a check should be made of any nonlinear ele-

ments within the test object, e.g. tapping protection or core earthing
as a source of the disturbance.
Voltage oscillograms
Applied voltage oscillograms are relatively insensitive and indicate
major faults in the insulation. However, the indication of the fault in
the following cases can be obtained from voltage oscillograms.
— Faults in a major insulation at or near the line end will result
in a rapid and total collapse of the voltage.
— A total flashover across the winding under test will result in
a slower voltage collapse, normally occurring in steps.
— A flashover across the part of the winding reduces the
impedance of the winding, thus resulting in a decrease of the
time to half value. Oscillations may also occur on the voltage
wave.
— Minor faults such as breakdown of coil-to-coil or even turn-to-
turn insulation may sometimes be detected as high frequency
oscillations on the voltage oscillogram.
Transferred voltage oscillograms also indicate these faults and the
sensitivity of measurement is higher than that of applied voltage.
Current oscillograms
Current oscillograms are the most sensitive means of failure detec-
tion. However, this sensitivity is accompanied by the possibility of
indicating apparently a number of defects, which may not directly be
associated with failure like sparking in the external earth circuit,
poor earth connection within the tank, core sparking, etc. Thus,
judgement and discrimination are required in the interpretation of
current oscillograms to avoid the possibility of a healthy transformer
being judged unsound.
Current oscillograms almost always verify and magnify the small
disturbances found on the voltage oscillograms. Major changes in
current oscillograms indicate probable breakdown within the wind-
ings and earth. Small, localized high frequency oscillations spread
over two or three microseconds are a possible indication of severe
discharges or partial breakdown in the insulation between turns or
coils or coil connections. An indication of even a short-circuited turn
is obtained in a current trace, as a change in shape or amplitude,
which may not be revealed in a voltage oscillogram. A short-circuit
means a decrease of the inductance, which results in a change in the

current through neutral shunt.
To have better judgement, current oscillograms are recorded
simultaneously on two time sweeps, a faster sweep and a slower
sweep.
(b) Switching Impulse

Voltage oscillograms
All types of faults normally cause a significant change in the voltage
wave, either as a temporary dip or more often as a complete collapse
of the wave or a shortening of the tail. The voltage records for a
switching impulse tests are a sufficiently sensitive means for detec-
tion of faults.
The wave tail shortening due to fault in general, is clearly distin-
guishable from the variation in the wave tail length due to magneti-
zation of the core.
Current oscillograms
The current oscillograms taken on successive impulses may not
be comparable, since the voltage waves themselves may not be iden-
tical. Current oscillograms should be looked for burst of oscillations,
which would occur at approximately the same time as any distortion
in voltage wave.
REFERENCES
(For Section II)
1. IEC Standard 60076–3. Bureau Central de la Commission
Electrotechnique Internationale, Geneva, Switzerland.
2. IEC Standard 60060–2 and 60060–3. Bureau Central de la
Commission Electrotechnique, Internationale, Geneva, Switzerland.
3. IEEE Guide for Transformer Impulse Tests C57.98 IEEE No. 93
June 1968.
4. Feser, K., Problems Related to Switching Impulse Generation at High
Voltages in the Test Plant—Haefely Publication.
5. Sie, T.H., Switching Surge Tests of Power Transformers—Bulletin
Oerlikon No. 383/384.
6. Switching Surge Tests For Oil Insulated Power Transformers—
IEEE Committee Report, IEEE Trans., PAS 87 No. 2, 1968.
SECTION III
14.3 Partial Discharge Testing
Partial discharge test has recently been introduced as a routine test

for transformers and reactors designed for system voltages ≥ 300 kV.
This is basically a long duration test intended to check the insulation
with regard to voltage under normal operating conditions and tem-
porary over-voltages, generally originating from switching operations
and faults (e.g. load rejection, single-phase faults). The magnitude of
these over-voltages do not exceed 1.5 p.u.(4). The test proposed in the
standards is based on the assumption that a voltage of 1.5 times the
rated voltage applied for 30 min is suitable for checking the ability of
the insulation to withstand the stresses mentioned above by judging
its electrical discharge behaviour.
14.3.1 Definition of P.D.
A partial discharge can be defined as localized electrical discharge in

insulating media which only partially bridges the insulation between
conductors.
Some of the conditions that can initiate partial discharges in
transformers are:
— Improper processing or drying of the insulation.
— Over stressed insulation due to a lack of proper recognition of
the voltage limitation of the insulation.
— High stress areas in conducting parts, which can be caused by
sharp edges on either the conducting part or the ground
plane.
These factors may cause ionization in cavities within solid insula-
tion, in gas bubbles in insulating liquids, or along dielectric surfaces.
Although, involving small amounts of energy, the partial discharges
may lead to progressive deterioration of the dielectric properties of
insulating materials.
14.3.2 Theory of P.D. Measurement
Consider Fig. 14.11, the ideal case where a single cavity exists within
the dielectric. The cavity can be compared to an air condenser C2.
A A
C2
U Cp
C1
B B
Figure 14.11 Dielectric with a cavity.
The remainder of the column of the dielectric having same diameter

as the cavity is represented by a perfect condenser C1 in series with
C2. The remainder of the dielectric forming a perfect condenser is
represented by Cp.
When a test voltage U is applied, distribution within the insula-
tion will assume its steady-state value u1(t) and u2(t) across C1 and C2
respectively, according to the relative capacitive coupling conditions
If u1(t) across C1 approaches the break down voltage Uz, then the
dissipated charge within C1 will be
q1E = C1 ◊ Uz
While C1 is bridged by the discharge of q1E, Cp will transfer a
charge into C2 through C1 until charge equilibrium is attained
between C2 and Cp. This charge transfer will cause a voltage drop
across the terminals A–B. The charge dissipated within C1 will then
be
q = Cp ◊ D U = C2 (Uz – D U)
for Cp >> C1, D U will be << Uz, hence
q = C2 ¥ Uz
The total charge dissipated within C1 is then given by

q1 = q + q1E
This charge, however, cannot be measured, because it does not
pass through connections external to the insulation and further, the
values of C1 and C2 are not known. The charge, accessible in this
condition, is the apparent transferred charge of the insulation, which
is determinable by measurement across the terminals A–B. Accord-
ing of IEC Publication 270, the apparent charge is defined as follows.
“The apparent charge ‘q’ of partial discharge is that, which, if in-
jected instantaneously between the terminals of the test object, would
momentarily change the voltage between its terminals by the same
amount as the partial discharge itself ”.
14.3.3 Principle of Measurement
The absolute value |q| of the apparent charge is referred to as the

discharge magnitude and is expressed in coulombs. Figure 14.12
shows a basic partial discharge test circuit.
Ck
C
Ca
U P.D. meter
Zm
D
B
Ck = coupling capacitor capacitance Ca = test object capacitance
Figure 14.12 Basic partial discharge test circuit.
On the inception of P.D. in the test object, the voltage across the
terminals A = B will be U – DU at the instant t = 0, because of the
voltage drop D U caused by the charge transfer. However, across the
capacitor, the voltage at the instant t = 0 will still be U, due to the

time constant introduced by the series resistance Zm (measuring im-
pedance).
The charge transferred from Ck via Zm into Ca, i.e. the measured
charge will be
• •
DU DU
qm = z
0
i(t) dt =
Zm z
0
e - t /t ◊ dt =
Zm
◊t
with
t = Zm
FG C ◊C IJ
k a
HC + C K
k a
the above equation becomes
qm = DU ◊ Ca Ck
Ck + Ca
1
=q◊
FG C
1+ a
IJ
H Ck K
\ The apparent charge
q = qm 1 +
FG Ca IJ = q
m ◊K
H Ck K
The factor K, by which the measured charge has to be multiplied
to obtain the apparent charge, is called the correction factor. It is
observed that this factor is affected by the circuit characteristics,
especially by the ratio of the test object capacitance to that of the
coupling capacitor.
For each new test object, the correction factor is determined by
calibration of the P.D. instrument with the test object connected in
the circuit. A calibration charge of known magnitude q0 is injected
between the terminals A–B and the calibration charge q0m indicated
on the P.D. measuring instrument is registered.
q0
The correction factor is then calculated as K = .
q 0m
14.3.4 Test Circuits

The three basic types of test circuits for measurement of partial dis-
charges are described in IEC 60070(2). Each of these circuits mainly
consist of
— a test object, which in general can be regraded as a capacitor,

Ca
— a coupling capacitor, Ck
— the measuring circuit consisting of a measuring impedance,
connecting leads and the measuring instrument (generally
some wide band instruments like ERA Model III or V and
Nonius differential detector according to Dr. Kreuger).
For high voltage transformers, reactors, etc., usually, the mea-
suring circuit is connected between the bushing tap (when available),
and the tank (Figs. 14.13 and 14.14) so that the bushing capacitance
acts as the coupling capacitance to the HV terminal. It is sometimes
HV
N
L
Zm
Discharge
detector ~
LV
(PDD)
Figure 14.13 Test circuit for P.D. measurement on transformer at

bushing test tap.
necessary to connect an adjustable inductance across the measuring

terminals or a matching reactance network, giving maximum
response over the measuring frequency band.
14.3.5 P.D. Measurement
Typical circuit arrangement for P.D. test is shown in Fig. 14.15. The
connections shown are for the HV line terminal and the neutral
HV
IV
L L
Zm Zm
PDD PDD
~
LV
Figure 14.14 Test circuit for P.D. measurement at two terminals.
terminal, either during three-phase of single-phase excitation. Fig-

ure 14.15(a) and (b) show three-phase and singe-phase supply ar-
rangements respectively. The preferred arrangement is to test
three-phase transformers by single-phase excitation, so that the dis-
charges are confined to the phase under measurement.
Measurement by multi-terminal method consists in calibrating
the complete measuring circuit and measuring the discharges dur-
ing excitation of the transformer at the required test voltage. The
measurements are made at neutral M1 and at the bushing tap M2 of
the line terminal as shown in Fig. 14.15(c). Measurements at other
terminals of the transformer are also done similarly.
14.3.6 Duration of Test and P.D. Values
Figure 14.16 indicates the time sequence of the test voltage, as per
IS : 2026.
The test voltages between line and neutral terminals are ex-
pressed in P.U. of Um/ 3 as follows:
U1 = 3 Um/ 3 = Um, the highest voltage for equipment

U2 = 1.5 Um/ 3 or 1.3 Um/ 3
LV LV
N N
HV HV
Um Um Um Um 0.5 Um 0.5 Um
3 Um 3 Um 1.5 Um
(a) (b)
1W 1V 1U Matching unit
Matching unit
2W 2V 2U M1 M2
PDD
(c)
Figure 14.15 Typical test circuit arrangements.

U
5 sec
U1
5 min 30 min
U2
Figure 14.16 Time sequence for the application of test voltage.
Throughout the application of test voltage, the value of apparent

charge should not exceed the values specified in standards, i.e. at 1.3
Um/ 3 : 300 pC and at 1.5 Um/ 3 : 500 pC.
14.3.7 Disturbances and Measurement Precautions

Disturbances are the electrical discharges which do not come from
the test object.
The principal forms of disturbances are as follows:
— Interference from the supply mains.
— Interference from the earthing system.
— Pick-up from other high-voltage test or electromagnetic
radiation.
— Discharges in the test circuit.
— Contact noise caused by bad contacts in the circuit.
For the correct interpretation of the partial discharge behaviour
of the test object, it is necessary to ensure that the origin of the
measured partial discharges lies in the test object and not in the
measuring circuit or the surroundings. To know the disturbances
from the measuring circuit, the measurement is taken before apply-
ing the voltage. If necessary, proper corrective action is taken to
limit the discharge level well below the maximum permissible value
for the object under test. The instrument deflection is considered
only when the discharge from the test object exceeds the ambient
level. When the voltage is raised, the unearthed metallic objects ly-
ing in the surrounding area can also cause disturbances, as they
behave as floating potential points. It is, therefore, desirable that as
far as possible, the surrounding area should be free from loosely-
lying metallic parts which are not necessary for the test.
The disturbances caused by electromagnetic fields of rotary ma-
chines, rectifiers, radio transmitters, etc., can be effectively checked
by proper shielding of the laboratory, which acts as a Faraday cage.
This also proves to be a very effective measure for carrying out P.D.
measurements, without the need for often very lengthy and trouble-
some search for causes of disturbances.
Above all, the high voltage connections also should be made dis-
charge free, and the sharp points should be properly shielded using
corona control rings or shields.
14.3.8 Location of P.D.
Precise location of P.D. requires great skill and experience. Several

methods are in vogue for the location of partial discharges. Some of
them are described here briefly.
(a) Acoustic Detection

For the location of corona discharges in air, directionally selective
microphones with high sensitivity above the audible frequency range
are used.
(b) Visual Detection

Visual observations are carried out in a dark room with or without
the aid of field-glasses of large aperture, Alternatively, a photo-
graphic record can be made, but fairly long exposure time is usually
necessary. For special purposes, photomultipliers are sometimes
used.
(c) Electrical Location

The method consists in multi-terminal calibration by injecting
known charge between each pair of terminals available (e.g. HV ter-
minal and tank, neutral terminal and tank, HV and neutral termi-
nal, HV and bushing tap). This procedure makes it possible to know
the instrument response at different measuring terminals for each
simulated P.D. position. A correct evaluation of the P.D. source loca-

tion can be done by comparing the ratio between two simultaneous
readings at different terminals during the HV test with the corre-
sponding ratio obtained during circuit calibration. The closest pair of
ratios give the approximate electrical location of the P.D. source.
(d) Ultrasonic Method

The methods described above do not enable geometrical location of
the source of P.D. For example, a P.D. source, electrically located
close to the middle terminal of an auto-connected transformer may
be at any place along the connecting leads, between the series and
common windings, or at the adjacent winding ends. It, therefore,
becomes difficult to ascertain whether the discharges are harmless
or they need to be eliminated.
Ultrasonic method facilitates location of the space coordinates of
the P.D. source with respect to the transformer tank.
The principle of ultrasonic method of location is very simple.
Partial discharges produce high frequency electrical pulses that
spread through the windings and can be picked up at the bushing tap
almost instantaneously.
Partial discharges, being very similar to small explosions, also
produce pressure waves that are propagated in all directions and can
be collected by the transducers placed on the outer surface of the
tank as shown in Fig. 14.17. Figure 14.18(a) and (b) (Plate 6) show
the photographs of instruments connected for the measurement and
location of P.D.
Each transducer picks up the ultrasonic signal with delays,
depending on its distance from the P.D. source. The time of propaga-
tion of the pressure wave may be assumed as equal to the delay
between electric pulse at the terminals and the signal detected by
the transducer.
The delays obtained with a few transducers make it possible to
trace the origin of the partial discharge by a trigonometric proce-
dure. In principle, the delay of the ultrasonic signal can be observed
with an oscilloscope, triggered by the terminal electric pulses. But in
practice, such a procedure is difficult, since the ultrasonic signal is
mixed with random and spurious signals due to a high partial dis-
charge repetition rate, background noise and transducer vibrations.
The electrical pulses appear at time intervals shorter than those
Ultrasonic signals
Tank wall
Transducer
Ultrasonic Magnetic
detector tape
unit recorder
Discharge source
Electrical
signal
P.D. Trigger Signal

meter signal averager
Input
Figure 14.17 Location technique of discharge source.
necessary for ultrasonic wave propagation and transducer damping.

Thus the ultrasonic signal at the transducers is practically continu-
ous (Fig. 14.19).
Moreover, if the discharges have a low amplitude (i.e. low signal-
to-noise ratio), the ultrasonic signal may be masked by background
noise.
This problem can be overcome by using a signal averager instru-
ment for recording the ultrasonic signals. As the noise is generally
of random nature, the summation averaging procedure of signal
averager results in signal-to-noise ratio improvement, proportional
to the square-root of the number of summations. However, high
frequency disturbances may interfere with the electric pulse used for
triggering.
14.3.9 Characteristics of Partial Discharges
The characteristics of partial discharges are shown in oscillograms

in Fig. 14.20, according to their origin, i.e. partial discharges of a
metallic point, partial discharge caused by several metal parts with
free potential against each other and partial discharges caused by
bad contacts.
3
2
(a)
20 mm sec.
3
(b)
(c)
Figure 14.19 Ultrasonic signal masking due to PD high

repetition rate.
(a) test voltage and high frequency pulses
(b) ultrasonic signal produced by single P.D. and
relevant delays
(c) their super position at the transducer.
PC PC
U U
t
t
PC
PC
U U
t
t
Partial discharges of a Partial discharges caused by

metallic point several metal parts with free potential
PC
PC
U U
t
t
Bad contacts
PC
U
t
Partial discharges caused by

a metallic part with free potential
Figure 14.20 Partial discharge superimposed on AC voltage.

REFERENCES
(For Section III)
1. Specification for Power Transformers IS : 2026 (Part III)—1977.
2. Partial Discharge Measurements, IEC Publication 60270.
3. Methods for PD Measurement, IS : 6209—1998.
4. Nieschwietz, H. and Stein, W. Partial Discharge Measurements as
Means of Quality Control of High-Voltage Power Transformers,
Conference on Partial Discharge in Electrical Insulation, April 1976,
Bangalore.
5. The Prachauser: Locating Partial Discharges in High Voltage
Equipment, ibid.
6. “Measurement of Partial Discharges in Transformers” Electra,
No. 19, Nov. 1971, pp 13-65.
7. Recognition of Discharges—Electra, No. 11, pp 61-98.
8. Mora, P. and J. Poittevin, Location of Partial Discharges in the
Transformers to be Detected with the Help of Ultrasonic Waves.
SECTION IV
14.4 Testing of Reactors
The fundamental difference between a transformer and a shunt rec-

tor is that, the latter cannot be tested at no-load, like a transformer.
The reactor will draw rated power when rated voltage is applied.
Large and expensive test plant is needed for carrying out most of
the tests on shunt reactors, especially on large EHV units. Power
capacities of the factory test laboratories have in the past, normally
been decided by the requirements for power transformer tests, viz.
load-loss test and heat-run test. When EHV shunt reactors of large
size were designed and manufactured for the first time two decades
ago, the test sources were insufficient even for testing the reactors
up to rated voltage for short time. This was the situation for large
low-voltage reactors also.
Manufacturers of large EHV transformers have, during recent
years, increased their test facilities for transformers. This has at the
same time increased the resources for reactor testing and special
facilities for such tests have also been added.
Recommendations of IEC Publication 60289 Reactors, apply to the
following types of reactors
— series reactors, such as current limiting reactors and load
sharing reactors;
— shunt reactors;
— single-phase neutral earthing reactors and arc-suppression
coils;
— three-phase neutral electromagnetic couplers and earthing
transformers.
Because of the similarity between power transformers and
the reactors, many of the provisions of IEC : 60076 ‘Power Trans-
formers’ apply also to these reactors. Tests on reactors comprise the
following:
(a) Routine Tests

(i) Measurement of winding resistance
(ii) Measurement of reactance
(iii)Measurement of loss
(iv) Inter-turn over voltage withstand test
(v) Separate source voltage withstand test
(vi) Measurement of insulation resistance of windings
(vii) Partial discharge measurement (as applicable)
(viii) Lightning impulse voltage withstand test (routine and type
test)
(ix) Switching impulse voltage withstand test (as applicable).
(b) Type Tests

In addition to the routine tests, the following type tests are con-
ducted, if specified by the purchaser
(i) Temperature rise test
(ii) Lightning impulse voltage withstand test
(c) Special Tests

Following special tests are conducted on the reactors where appli-
cable
(i) Measurement of zero-sequence impedance
(ii) Measurement of vibration and acoustic sound level
(iii) Measurement of third harmonics in phase currents
(iv) Lightning impulse voltage withstand test on neutral
(v) Short-circuit current withstand test—-applicable to series
reactors and earthing transformers only.
While majority of the tests applicable to different types of reactors
are similar in details, a few tests are the particular requirements for
particular reactors.
As large EHV shunt reactors pose some problems in testing, the
tests generally applicable to shunt reactors have been discussed
in detail. Wherever necessary, special notes have been added to
bring out important differences in requirements for other types of
reactors.
14.4.1 Measurement of Winding Resistance (Routine Test)
This is similar to testing of power transformers and covered in Sec. I.

14.4.2 Measurement of Loss (Routine Test)
The measurement of loss is carried out preferably at rated voltage

and frequency. Should this be impracticable, the measurement
method is subject to agreement between manufacturer and pur-
chaser on the understanding that the test voltage chosen is as high
as possible, and that tests at different voltages are conducted for the
purpose of extrapolation.
In case of series reactors, impedance is a guaranteed figure. It is
measured at rated frequency by applying a voltage, which results in
the circulation of the continuous rated current. The ratio of the ap-
plied voltage to the rated current gives the impedance value, e.g. in
case of three-phase series reactor, the impedance value is taken as
Phase-to-phase applied voltage
Continuous rated current ¥ 3
Unlike transformers, the loss in the various parts of the reactor
cannot be separated by measurement. It is thus preferred that in
order to avoid correction to reference temperature, measurement is
performed when the average temperature of the windings is practi-
cally equal to the reference temperature. If this is impracticable, the
loss may be broken down into core-loss, I2R-loss and stray loss and
the individual components corrected to rated voltage and to the
reference temperature in accordance with the requirements of
standards. The core-loss is taken to be the calculated core-loss and is
deemed to be independent of temperature. The I2R component of the
measured loss is calculated from the measured resistance of the
winding. The remaining loss is considered to be the stray loss.
If from the results of type tests on a similar reactor, the losses
both at reference temperature and in the cold state are known, the
loss on the reactor under routine test can be corrected to reference
temperature as follows:
If Wr is the cold-state loss of routine tested reactor at temperature t1,
If Wt is the cold-state loss of type tested reactor at temperature t2,
If Wft loss of type tested reactor at reference temperature,
If Wfr loss of routine tested reactor at reference temperature (to be
determined),
then
Wf t 235 + t2
Wfr = Wr ¥
Wt 235 + t1
14.4.3 Inter-turn Over-voltage Withstand Test (Routine Test)
Because of large power requirements, this test is generally impracti-

cable even with increased frequency for large reactors, This test,
when impracticable, is replaced by an impulse-voltage test. Details of
impulse test are similar to those applicable for transformers and are
covered in Sec. II.
14.4.4 Separate Source Voltage Withstand Test (Routine Test)
The reactor is tested at the specified test level. The test voltage is
applied between the line and neutral terminals joined together and
the earth for 60 s.
14.4.5 Measurement of Insulation Resistance (Routine Test)
The insulation resistance of the winding with respect to core and

end-frame and tank connected together to earth is measured. The oil
temperature is measured immediately prior to the test and recorded.
14.4.6 Partial Discharge Measurement
In order to check the level of internal partial discharges at specified

over-voltage, partial discharge measurement is performed on all
EHV reactors of 300 kV and above class, as a routine test. The meth-
odology of conducting the test is similar to that applicable for power
transformers and is covered under Sec. III.
14.4.7 Lightning Impulse Voltage Withstand Test

(Routine and Type Tests)
Impulse test is intended to check the ability of a reactor to withstand

steep-wave impulse-voltages between turns of the windings. Details
of the test with full wave and chopped wave are covered under
Sec. II.
14.4.8 Switching Impulse Voltage Withstand Test
The test in intended to check the internal main insulation to earth of

the reactor against switching transient over-voltages caused by
switching operations in the systems.
This is a routine test for windings with highest voltage ≥ 300 kV.
The method of testing is similar as applicable for power transformers
and is covered under Sec II.
14.4.9 Temperature Rise Test (Type Test)
This test is performed at rated voltage and frequency to demonstrate

the thermal performance of the reactor. When the capacity of test
plant does not permit the test to be performed at rated voltage, it can
be done, subject to agreement between manufacturer and purchaser,
at a voltage as close as prescribed to the rated voltage and the results
are corrected to rated conditions in accordance with IEC : 60076.
14.4.10 Measurement of Zero-Sequence Impedance
This test is not applicable to series reactors. For star-connected

three-phase shunt reactors, the zero-sequence impedance (in ohms
per phase at rated frequency) is measured by applying rated phase-
to-neutral voltage between the line terminals connected together
and the neutral.
14.4.11 Measurement of Vibration and Noise Level

(a) Vibration Measurement
The vibration measurements are taken to check the safety of the
reactor tank against any impending mechanical damage.
The vibrations are generated due to the magnetostrictive strain
in the core and can magnify due to resonance of the core itself. Local
resonance of auxiliary components may also magnify the vibrations.
For conducting vibration measurements, the reactor is energized
at the specified voltage and at the rated frequency. An accelerometer
is fixed at the location where measurement is intended. The vibra-
tion signal from accelerometer is transmitted to a vibration meter
which indicates the vibration level in microns.
(b) Acoustic Sound Level Measurement

In order to ascertain the noise intensity, sound level measurement
is made during factory tests. The main purpose of this measurement
is to verify that the design has not hit mechanical resonance, either
for the active parts or for the different parts of the tank.
The vibration of the iron circuit is the major cause of noise gener-
ated in reactors. The noise level depends upon the flux density and
the magnetostrictive strain.
Sound level measurements are taken in accordance with NEMA-
TR1 1971. The reactor is energized at the specified voltage at rated
frequency. The measurements are taken at approximately the
centre height of the reactor tank, if its height is less that 8 ft and at
approximately one-third and two-third heights of the tank, if its
height is 8 ft or more. The measurements are taken around the reac-
tor at intervals of 3 ft. The microphone is kept at a distance of 1 ft
from the main noise emitting surface. In case the radiator fans are in
operation, the microphone is kept at a distance of 6 ft from any
portion of the radiators or coolers. The correction for background
noise, if any, is applied. The noise level of a reactor is the arithmetic
average of all the readings obtained at different locations with appro-
priate correction for background, if applicable.
14.4.12 Measurement of Third Harmonics
The third harmonics in the phase currents add together at the

earthed neutral. Consequently the earth current can be high, lead-
ing to disturbances. It is a normal requirement that the total har-
monic content in the phase current of a star-connected HV reactor,
and particularly the zero-sequence component appearing in the
neutral shall stay below a specified limit, which is of the order of
1–3% of rated current.
In order to ascertain the magnitude of third harmonic current,
the reactor is energized at rated voltage and frequency with the neu-
tral solidly grounded and measurements done. The verification by
measurements is sometimes difficult because the harmonics in the
energizing voltage are of comparable order, to the contribution from
reactor reactance non-linearity.
3-phase supply
2 4 3
1. Shunt reactor under test

2. Current transformer
3. Harmonic wave analyzer
4. Shunt
Figure 14.21 Scheme for measurement of current harmonics in a

shunt reactor.
14.4.13 Lightning Impulse Voltage Withstand Test

on a Neutral Terminal
When the neutral terminal of winding has been designed for impulse
withstand voltage, it is verified by applying an impulse voltage
through any one of the line terminals or through all three line termi-
nals of a three-phase winding connected together. The neutral ter-
minal is connected to earth through an impedance and the voltage
amplitude developed across this impedance, when the impulse
applied to the line terminal, is equal to the rated withstand voltage of
the neutral terminal.
Alternatively, an impulse test corresponding to the rated with-
stand voltage of the neutral is applied directly to the neutral, with all
line terminals earthed. In this case, a longer duration of the front
time is allowed, up to 13 microseconds.
14.4.14 Short-circuit Current Withstand Test

(Special Test on Series Reactors)
This test is designed to prove the mechanical capability at rated

short-time current, and in the case of a reactor with a magnetic core
or shield, to measure the impedance corresponding to the rated

short-time current.
The test consists of two applications of rated short-time current,
each of a duration of ten cycles, the first peak having a crest value of
1.8 ¥ 2 times the rms value. (Certain service conditions may result
in asymmetry factors less than 1.8 ¥ 2 ).
Three-phase reactors are subjected to two three-phase tests—one
with maximum asymmetry of the short-circuit current in the centre
phase, the other with maximum asymmetry in an outer phase.
The proof of the reactor having withstood the test satisfactorily, is
determined by remeasuring the impedance; by visual inspection and
by repetition of the routine dielectric tests at 75% of the initial value.
14.4.15 Specialities of Reactor Testing
Requirement of large reactive power is sometimes a constraint for

testing of large EHV reactors at rated voltage. If the test facilities
permit, the reactor should be energized up to the rated voltage and
routine tests be conducted.
The difficulties in testing can mainly be due to
— Inadequate power capacity of test plants:
— Lack of a reliable and straightforward method of loss
measurement. In the absence of International consensus, IEC
recommends two alternative methods apart from the conven-
tional wattmeter method, which is subjected to some errors.
— Limitations imposed by factory conditions affecting measure-
ment accuracies.
The power frequency over-voltage tests and the loss tests have been
described hereafter with special reference to the specialities in-
volved in testing.
14.4.16 Over-voltage Tests
A check of level of internal partial discharges at specified test voltage

is becoming a generally agreed condition for acceptance of EHV
transformers. Application of an over-voltage test for large shunt re-
actor, corresponding to the induced voltage test for a transformer, is
a major difficulty, because of the prohibitive reactive power demand,
even for short duration testing. Testing at elevated frequency is the

only practical possibility (as for transformers) and is examined here
by elementary calculations.
Symbols used:
UN — rated voltage U1 — U|
fN — rated frequency f1 — |V Corresponding test
XN — rated impedance X1 — || quantities
QN — rated reactive load Q1 — W
K — ratio of test quantity and rated quantity:
Subscript defines the quantity under consideration.
The voltage and frequency used for the test are:
U1 = K u ◊ UN ; K u > 1 (1)
f1 = Kf ◊ fN ; Kf > 1
Reactor impedance
X1 = XN ◊ Kf
The current and flux density at any point:
I1 = IN ◊ Ka ◊ (Kf)–1 (2)
B1 = BN ◊ Ku ◊ (Kf)–1 (3)
The reactive load:
Q1 = QN ◊ (Ku)2 (Kf)–1 (4)
2
I R loss vary with the square of the current. Eddy losses in the
winding and in metal parts are proportional to the square of fre-
quency and to a varying power of the flux density. In the winding
with stranded conductors, it is assumed proportional to B2, but for
solid parts with pronounced skin effect, the power tends toward B0.5.
Losses in the winding dominate, however, eddy losses in the winding
are reduced to small proportion of the I2R losses. The total winding
losses are
P1 = (I2R)N ◊ (Ku)2 ◊ (Kf)–2 + PeN ◊ (Ku)2 (5)
It is to be noted that eddy losses Pe in the winding are independent
of frequency as long as the test voltage in constant.
Example 14.1 For inter-turn over-voltage withstand test on EHV

reactors with
Ku = 2.0, Kf = 4, PeN = 0.12 (I2R)N
Winding losses at rated conditions, and during the test will be:
PN = (I2R)N ◊ 1.12
P1 = (I2R)N ◊ (2.0)2 ◊ (4)–2 + 0.12 (I2R)N ◊ (2.0)2

= 0.73 ◊ (I2R)N
The reactive power demand will be
Q1 = QN ◊ (2.0)2 ◊ (4)–1
= QN ◊ 1
Thus it is observed that the reactive power consumption at 200
Hz is equal to the rated power.
Example 14.2 For partial discharge test with

Ku = 1.5, Kf = 4, PeN = 0.12 (I2R)N
Winding losses at rated conditions, and during the test will be:
PN = (I2R)N ◊ 1.12;
P1 = (I2R)N ◊ (1.5)2 ◊ (4)–2 + 0.12(I2R)N ◊ (1.5)2
= (I2R)N 0.41 = 0.37 ◊ PN
The reactive power demand is
Q1 = QN ◊ (1.5)2 ◊ (4)–1 = 0.56 ◊ QN
It is observed that there is reduction in active and reactive
power consumption. However, the effect of these conditions on
local hot-spots should be examined.
In a shell-form winding, first, the local flux pattern is very regu-

lar, varying from a maximum value at the inner edge to zero at the
outer edge. This variation means that local eddy losses at the inner
edge are three times as high as that indicated by the average for
the whole winding. A comparison between rated conditions and test
conditions analogous to Eq. (5) above, thus gives:
PN (hot-spot) ~ (I2R)N + 3 ¥ 0.12 (I2R)N ~- 1.36 ◊ (I2R)N
–2
P1 (hot-spot) ~ (I R)N ◊ (1.5) ◊ (4) + 0.36 (I2R)N ◊ (1.5)2
2 2
~ 0.95 ◊ (I2R)N
with Ku = 2.0
P1 (hot-spot) ~ 1.69 ◊ (I2R)N
It remains to be checked, how the static shields, if used, would
behave. Their losses would go up by the square of test voltage,
independent of frequency.
In a gapped core reactor, the same basic relations apply, but the
distribution of fringing flux around the air gaps and consequently,
the local winding eddy losses are greatly dependent on design.
It may be concluded that

— The inter-turn over-voltage test, the test level for which is
double the rated voltage, is practicable at an elevated
frequency of 200 Hz, but may cause thermal injury to
insulation.
— The alternative test recommended to replace this test is the
full-wave lightning impulse test. Although the lightning
impulse test is more onerous, because of high voltage
gradients in a few initial turns resulting from non-linear
impulse voltage distribution along the winding, yet this test
is the most appropriate substitute test.
The partial discharge test, which is recommended at 1.5 times the
rated voltage, can be conducted with advantage at an elevated fre-
quency. This precludes the possibility of any risk of thermal injury to
the reactor under test.
14.4.17 Loss Measurement
It is well known that the difficulty associated with loss measurement

on reactors lies in determining these with sufficient accuracies at
power factors less than 0.004. The common methods of loss determi-
nation in practical use are:
(a) Calorimetric measurement with oil-to-water heat exchanger.
(b) Bridge method giving the loss angle of reactor impedance.
(c) Direct wattmeter measurement.
It is generally agreed that the calorimetric test can be considered
as the reference test method, which gives good accuracy when car-
ried out with care. It is virtually unjustifiable as a routine test due to
considerable cost and time involved in measurements. Normally
Bridge method is used for loss measurement of reactors however
periodic calibration of the bridge is recommended with calorimetric
method. Wattmeter method is the most straight-forward method
when accuracy requirement is not extremely high.
(a) Calorimetric Loss Measurement

The principle of calorimetric test is to determine the power carried
away in the cooling medium by measuring its rate of flow and
its temperature rise. In water-cooled apparatus with closed cooling
systems, measurements can be performed either in air or oil or in

the water circuit. In view of the difficulties in making measure-
ments, where air is the cooling medium, and uncertainty as regards
the specific heat of oil, the calorimetric measurements are carried
out predominately on the water circuit. To the power carried away
by the cooling medium, the heat lost by virtue of radiation and con-
vection, or by other means not assessed in the cooler, is added.
The reactor is energized at rated voltage and frequency. The ex-
ternal surface of the reactor, including the pipe work, is thermally
insulated by glass-wool to minimize the surface radiation. Oil is cir-
culated in the winding by a pump. The losses generated in the reac-
tor, after thermal equilibrium has been reached, are transferred
from the reactor to the exterior, either in the cooling water or by
convection and radiation from the surfaces of the tank, piping and
cooling equipment. Temperatures of oil and water are measured at
the inlet and outlet of the cooler. Water flow is measured by a water-
flow meter. When steady-state is reached, winding temperature is
determined by resistance cooling curve, and power consumption by
the pump is measured. Losses of the reactor at the test temperature
at steady-state condition comprise:
— Losses evacuated by the water circuit.
— Losses through the lagging insulation.
— Losses of the pump motor.
Figure 14.22 shows a general circuit arrangement for loss mea-
surement on a three-phase shunt reactor by calorimetric method.
Cooling arrangement. As per the standards, the loss guarantees are
at an average winding temperature of 75°C. The calorimetric test
must therefore be so controlled that thermal equilibrium is reached
when the winding is at or near this temperature. In the case of a
water-cooled reactor, the reactor’s normal cooler is used for the calo-
rimetric test even though difficulty can sometimes be encountered
in reaching 75°C in the windings, because of low cooling-water
temperature. If the reactor is normally air-cooled, a separate oil
water-cooler is fitted in place of the air-coolers.
Accuracy of the method. A rigorous analysis of the expected probable
errors of calorimetric test, in general, if hardly possible in view of
many unexpected errors during the test. However, when measure-
ments are made with all precautions, the accuracy of the calorimetric
method can be better than that achieved by a conventional method,
3-phase supply
2 1
5 3 4 7
6
1. Oil inlet 2. Oil outlet

3. Water inlet 4. Water outlet
5. Water flow indicator 6. Oil-to-water heat exchanger
7. Discharge water collecting tank
Figure 14.22 Arrangement for loss measurement on a three-phase

shunt reactor by calorimetric method.
viz. the wattmeter method, because this measurement is not

affected by the reactive power.
Though calorimetric method is more expensive and takes more
time to perform than conventional methods, yet in certain cases, e.g.
for the determination of losses of a large high voltage shunt reactor,
this method is the only reference method, which gives good accuracy
and can be employed to calibrate the bridge.
(b) Bridge Measurement

With bridge circuit, a voltage proportional to the reactor current but
exactly 90° out-of-phase with respect to this current, is induced in
the mutual inductance M. This proportional but out-of-phase voltage
is measured magnitude-wise and phase-wise relative to the terminal
voltage of the reactor by means of Schering bridge by comparing the

test object with an almost loss-free standard capacitor, with accu-
rately known capacity and dissipation factor. The bridge is balanced
by adjusting resistance and capacitance until perfect balance is
achieved.
Figure 14.23 shows the scheme of bridge method for loss mea-
surement on a shunt reactor. Figure 14.24 (Plate 7) shows a
photgraph of the test set-up on a 30 MVAr, 33 kV shunt reactor.
Factors which affect accuracy. The bridge circuit, incorporating a
mutual inductor in series with the high-voltage reactor, is a preci-
sion component, which can be calibrated for loss angle with high
accuracy, the compressed gas capacitor used for dielectric loss
measurement is also accurate. However, in practice, the connection
is complicated by the introduction of current transformer in cascade
with the mutual inductor.
In a circuit without current transformer, the mutual inductor
primary constitutes an addition to the impedance of the specimen,
and the modification of the reactor current is negligible. With a
3-phase supply
CN
1 2
R4
C4
3 4 5
1. Shunt reactor under test

2. Compressed gas capacitor
3. Precision current transformer
4. Mutual inductor
5. Schering bridge
Figure 14.23 Bridge circuit for power loss measurement.

current transformer added, however, the mutual inductor primary is

a loading on the current transformer, and the resulting phase
shift represents the greatest individual term of uncertainty in the
whole set-up. In order to bring this bridge circuit up to a satisfactory
level, it is necessary to have a mutual inductor designed for the full
reactor current.
Further, the bridge circuit arrangements, particularly the ones
used for the measurements on objects which are producing harmonic
waves, e.g. high voltage inductances with closed iron-cores or power
transformers under no-load are frequency sensitive and small fre-
quency fluctuations in the applied voltage tend to reduce measure-
ment precision. Therefore, a measuring voltage free from frequency
fluctuations is necessary.
The bridge method is very sensitive of magnetic stray fields and
the current transformer is especially sensitive.
However, bridge method has its own merits over the calorimetric
method.
— Compared to the calorimetric method, the use of bridge
affords considerable economy in time.
— Use of a high-voltage compressed gas capacitor as a standard
capacitor is an advantage because it is generally available in
most high-voltage testing laboratories.
REFERENCES
(For Section IV)
1. IEC Publication 289 (1989); Reactors.
2. IS : 5553 Part II—1989; Shunt reactors.
3. BS : 4944: 1973—Specification for Reactors, Arc-suppression Coils
and Earthing Transformers for Electric Power Systems.
4. NEMA Standard Publication No. TR 1—1971: Transformers, regu-
lators and reactors.
5. Sollergren B., Special Report for Group 12 (Transfomers); CIGRE, Vol.
1970.
6. Grundmark B., High Voltage Shunt Reactors—Trends in Design and
Testing. Report 12—03 Ibid.
7. De Bourg H. and others, Calorimetric Loss Measurement On
Alternators and Reactors, Report No. 119; CIGRE Vol. 1964.
SECTION V
14.5 Short-Circuit Testing of Power Transformers
14.5.1 Title the Short-Circuit Test Requirement:

It is given in IEC 60076-5: 2000 and IS2026 (Part I). For large power
transformers, the margins are quite high for thermal effect whereas
dynamic effect is quite complex and requires tedious calculations.
This section elucidates the testing requirements of short-circuit tests
and short-circuit duty of power transformers.
This test is a special test and the facilities required for the same
are generally not available with the manufacturers as very large
amount of power is required. There are only a few laboratories in the
world that are equipped with such large rating test plants. Depend-
ing upon the voltage class, size (MVA rating) and the impedance
voltage of the transformer, specific confirmation is to be obtained
from the reputed laboratories for conducting this test.
As per IEC 60076-5, power transformers have been categorised as
under:
Category I — Up to 2500 kVA
Category II — From 2501 kVA to 100 MVA
Category III — Above 100 MVA
The symmetrical short-circuit current is calculated using the
measured short-circuit impedance of the transformer plus the
system impedance. The short-circuit apparent power of the system
is generally specified by the utility.
14.5.2 Requirements Prior to Short-Circuit Test

The transformer should pass all routine tests as specified in
IEC60076-1. However, the Lightning Impulse test is not required to
be conducted prior to the short-circuit test. The routine test report
must be available prior to the short-circuit test. At the beginning of
short-circuit test, the average winding temperature shall be prefer-
ably between 10° C to 40° C.
14.5.3 Peak Value of Short-Circuit Current

The first peak of asymmetrical short-circuit current is given by
I� = I ¥ 2 ¥ k
where I = symmetrical short-circuit current
k = accounts for initial offset of test current
2 = conversion factor for rms to peak value;
k ¥ 2 is called peak factor and depends on X/R ratio
where X = sum of reactances of the transformer and system
(Xt + Xs) in ohms
R = sum of resistances of the transformer and system
(Rt + Rs) in ohms
If system impedance is taken into account and X/R ratio of the
system is not known, then X/R ratio of transformer is considered.
Value of factor k ¥ 2 for different X/R ratios as given in standards
are I:
X/R 1 1.5 2 3 4 5 6 8 10 14
k¥ 2 1.51 1.64 1.76 1.95 2.09 2.19 2.27 2.38 2.46 2.55
Normally, X/R ratio is >14 hence
(1) For Category I transformers, k ¥ 2 = 1.8 ¥ 2 , i.e. 2.55
(2) For Category II and Category III transformers, k ¥ 2 = 1.9 ¥
2 i.e. 2.69
14.5.4 Tolerance on Asymmetrical Peak and Symmetrical

RMS Value of Short-Circuit Current and duration
of Short-Circuit Test
As per IEC 60076-5 (2000), the tolerances allowed are:

On peak asymmetrical short-circuit current = ±5% of specified.
On symmetrical short-circuit current = ±10% of specified.
Duration for short-circuit test = 0.5 sec for category I with ±10% Tol.
= 0.25 sec for category II and III with ±10% Tol.
14.5.5 Short-Circuit Testing Procedure of Two Winding

Transformers
Two types of short-circuit test procedures can be followed:

(1) Pre-set Method
(2) Post-set Method
(1) Pre-set Method

In this method, one winding is short-circuited prior to application of
the voltage to the other winding in order to obtain the desired short-
circuit current in both windings. In a power transformer, since in
most cases LV is near to core, hence normally HV is supplied with
power and LV is pre-shorted. This is essential in order to avoid satu-
ration of magnetic core which could lead to excessive magnetising
current superimposed on short-circuit current during the first few
cycles.
(2) Post-set Method

One of the windings is applied with AC supply of desired value, and
on the other winding, post short-circuit is done through a synchro-
nous switch. Thus, there is very low magnetisation inrush current.
This method, however, requires a power three times as high as the
pre-set method.
14.5.6 Short-Circuit Test Shots and Connections
Short-circuit test requires 3 shots on each phase. In case of 1-phase

transformer, one shot each shall be on the maximum tap, normal
tap, and minimum tap respectively, while in the case of 3-phase
transformers, a total of 9 shots are applied, 3 on each phase. Thus, 3
shots shall be applied at the maximum tap on one of the outer
phases, 3 shots at normal tap on middle phase, and 3 shots at mini-
mum tap on the other outer phase.
Further, the short-circuit test connections are given below in the
preferential order:
(a) 3-phase Connection: For a 3-phase transformer, a 3-phase
supply is to be used as far as possible. (See Fig. 14.24)
Zs = system impedance
S = synchronous switch for post-set method or a
rigid connection for pre-set method.
(b) 1½ Phase Connection: In case it is not possible to meet re-
quirements of cl. 14.5.6 (a) for 3-phase (especially for category
II and III transformers), then 1½ phase connection may be
used with a single phase supply on HV side. The connection is
given in Fig. 14.25
(c) Single Phase Connection: In case of a single phase trans-
former, this connection is used. Similarly, if power supply is
Zs
S
S
Zs
Zs HV side (star) LV side (delta) S

Zs = System impedance.
S = Synchronous switch for post-set method
or a rigid connection for pre-set method.
Figure 14.24 3 Phase method for S.C. test of transformer.
Zs
1- phase
AC supply
Zs
S
Figure 14.25 1½ Phase method for S.C. test of transformer.
insufficient for conducting 3-phase or 1½ phase method in a

3-phase transformer, then single phase method is used (in
case star connected winding and neutral is available). The
connection is given in Fig. 14.26.
Zs S
1-phase
AC supply
Zs S
HV side (star) LV side (delta)
Figure 14.26 Single phase method for S.C. test of transformer.

14.5.7 Testing Procedure
— Calibration shots upto maximum 70% are applied to adjust

and check the proper function of test setup with respect to
moment of switching on, current setting.
— The routine test should be complete and the test report
should be available.
— Short-circuit is applied at 100% level.
— Oscillographic record of applied voltage, current, as well as of
secondary side should also be recorded for each shot.
— Reactance shall be measured after each shot, and reactance
variation shall be within ± 2% for core type transformers with
concentric coils of category I and II, and within ± 1% for
category III transformers.
— Condition of passing of short-circuit test:
(a) Results of short-circuit test and measurements and checks
performed do not reveal any condition of faults.
(b) Dielectric tests successfully conducted/repeated.
(c) Out of tank inspection does not reveal any defects like
displacement, shift of laminations, deformation of windings,
connection or support structure etc.
(d) No trace of electrical discharges are found.
(e) Reactance variation shall be measured after each shot, and
reactance variation shall be within ±2% for category I
and II transformers, and within ±1% for category III
transformers. In case of category III transformers, if
the reactance variation is in the range of 1 to 2%, the
acceptance is subject to mutual agreement between the
manufacturer and the utility. For this, a more detailed
inspection of the windings after disassembly needs to be
carried out and analysed to establish the cause of the
deviation.
14.5.8 Optional Diagnostic Measures
One way to detect changes in the transformer due to forces during

short-circuit of transformer is to measure the Frequency Response
Analysis (FRA) before and after the short-circuit test. This is done by
measurement of frequency response (voltage transfer function or

admittance), which can be used to compare the frequency of the
windings response before and after the short-circuit test. Mechanical
change and voltage independent defects, like winding shorts are
reflected in the finger print.
Addendum to Chapter 14
356
Test Results of a Few Typical Transformers
315 MVA, 3-phase 250 MVA, 3-phase 200 MVA, 1-phase 100 MVA, 3-phase
50 Hz auto- 50 Hz generator 50 Hz generator 50 Hz auto-
transformer ONAN/ transformer ONAN/ transformer transformer
ONAF/OFAF cooled ONAF/OFAF cooled OFWF cooled ONAN/ONAF cooled
1. No-Load Ratio at 400/220/33 kV 15.75/420 kV 21/400 kV 220/110/11 kV
Principal Tap 3
2. Connection Symbol YN, a0, d11 YN, d11 YN, d11 when connected YN, a0, d11
in 3-phase bank
3. Lightning Impulse
Transformers, 2/e
HV 1300 kVp 1300 kVp 1300 kVp 900 kVp

IV 950 kVp — — 450 kVp
LV 250 kVp 125 kVp 125 kVp 170 kVp
4. Switching Impulse
HV 1050 kVp 1050 kVp 1050 kVp —
—
5. Partial Discharge at
Um
1.5 20 pC 35 pC 65 pC —
3
(Contd.)
(Contd.) Addendum to Chapter 14
Test Results of a Few Typical Transformers
315 MVA, 3-phase 250 MVA, 3-phase 200 MVA, 1-phase 100 MVA, 3-phase
50 Hz auto- 50 Hz generator 50 Hz generator 50 Hz auto-
transformer ONAN/ transformer ONAN/ transformer transformer
ONAF/OFAF cooled ONAF/OFAF cooled OFWF cooled ONAN/ONAF cooled
6. Separate Source Voltage

withstand
HV/Earth 38 kV rms 38 kV rms 38 kV rms 38 kV rms
LV/Earth 95 kV rms 50 kV rms 50 kV rms 28 kV rms
7. No-load Loss 101.4 kW 128.3 kW 92.4 kW 30.69 kW
8. % No-Load Current 0.06 0.07 0.14 0.05
9. Load Loss 274.1 kW 620.5 kW 476.7 kW 187.7 kW
10. % Impedance
HV–IV 13.34 — — —
Testing of Transformers and Reactors
HV–LV 51.38 14.51 14.66 8.36

IV–LV 33.10 — — —
11. Top Oil Rise 28.3°C 36.3°C 13°C 30.5°C
12. Winding Rise 48.6°C 45.0°C 31°C 44.4°C
357
C HAPTER 15
Standards on Power
Transformers
V.K. Lakhiani
S.K. Mahajan
Standards are evolved to meet a generally recognized demand,

taking into account the interest of manufacturers and users and
fulfilling the needs of economy. Today, a product cannot even be
visualized without a standard. A standard is a useful guide in all
facets of a product—conception, design, manufacture, testing, instal-
lation, operation, maintenance, etc.
Indian Standard 2026 “Specification for Power Transformers”
is the governing standard on power transformers. This has been
revised in 1977 with a view to bring it in line with the 1976 revision
of IEC Publication 76 ‘Power Transformers’. It is now in five parts
and is an exhaustive standard on power transformers.
Besides IS: 2026, other related standards, in extensive use, are
concerning electrotechnical vocabulary, fittings and accessories,
on-load tapchanger and its application guide, loading guide, new
oil, maintenance of oil, selection, installation and maintenance of
transformers, etc. The scope of these standards is briefly touched
upon in this chapter.
New standards in the offing are those connected with bushing air
clearance, guide to testing and application guide on power transform-
ers. An introduction to these new standards has also been made.
Efforts in different forums have been made to come out with a
standard specification of the transformer. “Central Board of Irriga-
tion and Power (CBIP) Specification” and “Central Electricity Author-
ity (CEA) Report” are the offshoots of such concerted efforts. A brief
mention has been made about these specifications. More and more
adaptation of these standard specifications, which purport to supple-
ment IS: 2026 is in the interest of the transformer industry.
Standards on Power Transformers 359
15.1 First Revision of IS: 2026
15.1.1 State-of-the-Art
IS: 2026 was first published in 1962 and covered initially natural
cooled oil-immersed transformers. By subsequent amendments (a
total of 8 in number) forced cooled transformers were included, use
of synthetic liquids as coding medium was permitted and require-
ments for minimum windings were incorporated.
The first revision was brought out in 1977 with a view to bringing
it in line with the revision of IEC Publication 76—1967, which was
revised in 1976.
At present the requirements for power transformers are covered
in five parts as follows:
PART I : 1977 General (First Revision)
PART II : 1977 Temperature Rise (First Revision)
PART III : 1981 Insulation levels and dielectric tests
(Second Revision)
PART IV : 1977 Terminal markings, tappings and connec-
tions (First Revision)
PART V : 1994 Transformer/Reactor bushing minimum
electrical clearances in air.
The second revision to Part III published in 1981. This revision
was exactly in line with IEC 76-3 (1980). However since then IEC 76-3
has undergone changes and now available as IEC 60078-3 (2000) :
Insulation levels, dielectric tests and external clearances in air. In
this version of IEC 60076-3 following new major points have been
introduced compared to IEC 76-3 (1980)
— Long duration AC (ACLD) test as a special test for 72.5 < Um
£ 170 kV and Routine test for Um > 170 kV
— LI is routine test for Um > 72.5 kV
— SI is routine test for Um > 170 kV
— Short duration AC (ACSD) as routine test for Um £ 170 kV and
special test for um > 170 kV
— Short duration AC (ACSD) test with P.D. for Um > 72.5 kV
— Induced over voltage test/P.D. test are not called and instead
ACLD and ACSD tests have been introduced.
Part 4 of IEC 76 tappings and connections has been clubbed with

IEC 76-1 revised in 1993.
15.1.2 IS : 2026-1977 vis-a-vis IS : 2026-1962–Major Points

of Difference
A few major differences partwise are enumerated below:
Part I
(a) The kVA ratings have been extended to an unlimited range in
the same series.
(b) Dry-type transformers have been covered.
(c) Terminology is now covered by IS : 1885 (Part xxxviii)–1977
‘Electrotechnical Vocabulary : Transformers’ (first revision).
(d) The reference ambient temperatures have been revised.
Maximum ambiet air temperature and maximum daily
average ambient air temperature have been increased by 5°C
to 50°C and 40°C, respectively. Concept of maximum yearly
weighted average ambient temperature has been introduced
and a hot-spot temperature of 98°C over this weighted
temperature of 32°C has been recognized, thereby giving a
66°C hot-spot rise.
(e) The requirements with regard to ability to withstand short-
circuit have been revised. A test under the category of ‘Special
Test’ for assessing the dynamic ability of the transformer
to withstand short-circuit has been included. The thermal
ability to withstand short-circuit has to be demonstrated by
calculation and the duration of the symmetrical short-circuit
current to be used for the calculation of the thermal ability to
withstand short-circuits has been specified as 2 s.
In fact, IEC 60076 has a separate Part 5 on this subject,
but in case of IS : 2026 (1977) the same has been clubbed with
Part I.
(f) Tolerances on losses, impedance voltage, etc., have been
revised. No-load current is also made a subject of guarantee
with +20% tolerance. The component losses have +15%
tolerance, while the total losses have +10% tolerance.
Impedances at taps other than those at principal tapping have
more tolerance than applicable for principal tapping by a
percentage equal to half the difference in tapping factor

(percentage) between the principal tapping and the actual
tapping.
(g) Operation test on on-load tapchangers and auxiliary circuits
insulation tests have been included in the routine tests.
(h) Supplementary tests now termed ‘Special Tests’ include
besides zero sequence impedance; short-circuit test, measure-
ment of acoustic noise level, measurement of the harmonics
of the no-load current and measurement of the power taken
by the fans and oil pumps.
(i) Requirement for overfluxing for generator transformers has
been added. A generator transformer subjected to load-
rejection conditions shall be able to withstand 1.4 times the
rated voltage for 5 s at the transformer terminals to which the
generator is to connected.
Part II
(a) Identification of transformers according to cooling methods
has been lined up with the international practice of specifying
by 4 letters, e.g., ONAN, ONAF, OFAF, ODAF, OFWF, etc.
(b) Directed flow type cooling has been recognized and has 5°C
more permissible winding temperature rises.
(c) The temperature limits are the same as those in IEC 60076-
2 (1993) viz. 100°C for top oil and 105/110°C for winding,
depending upon whether non-directed oil circulation or
directed oil circulation is adopted. As deviation from IEC, 5°C
more rises have been permitted for water cooled trans-
formers.
(d) Temperature rise limits are now valid for all tappings. For
multi-winding transformers, the temperature rise of the top
oil refers to the specified loading combination, for which the
total losses are the highest. Individual winding temperature
rises shall be considered relative to that specified loading
combination, which is the most severe for the particular
winding under consideration.
(e) Total losses to be fed for temperature-rise test should not be
less than 80%. Similarly, rated current should also be not less
than 90%. Values of indices ‘x’ and ‘y’ for the correction of oil
rise and gradient have also been revised.
(f) Winding resistance at the instant of switching off the supply

is now to be determined only by extrapolation of the curve
resistance versus time. The earlier method of applying
correction factors based on watts per kilogram of copper is
discontinued. The graphical method, though consumes time,
is more reliable, because average oil temperature can be
obtained on the graph which can serve as a cross-check.
Further, the trend of the curve shows whether the resistance
readings were stabilized.
Part III
(a) Below 300 kV, insulation requirements and dielectric
withstand tests are the same as before. Lightning impulse
test continues to be the type test. The other dielectric tests
are separate source power-frequency voltage withstand test
(for line terminals in case of uniform insulation and for
neutral terminal in case of non-uniform insulation) and
induced over-voltage withstand test, both to be performed as
routine tests. During the induced over-voltage test, at least
twice the rated voltage should appear across the winding,
having uniform insulation but in case of winding with graded
insulation, the line terminals must receive rated voltage
according to related table and excessive voltage between line
terminals is avoided. Neutral raising is allowed if it is so
designed. An impulse test for the neutral terminal is a special
test if a rated impulse withstand voltage for the neutral
terminal has been specified. A longer duration of the front
time up to 13 ms is allowed for the impulse test on a neutral
terminal.
(b) For voltage class ≥ 300 kV, there are two alternative methods
for specifying and testing the dielectric requirements, viz.
method 1 and method 2.
Method 1 is the same as applicable to voltage class < 300
kV, non-uniform insulation, except that lightning impulse
test is a routine test.
Method 2 introduces newer tests like switching impulse
and partial discharge measurement. According to this method,
the verifying dielectric tests are:
(i) Lightning impulse test for the line terminals (routine
tests)
(ii) Switching impulse test for the line terminals (routine test)
(iii) Separate source power-frequency voltage withstand test
for the neutral (routine test)
(iv) Induced power-frequency voltage withstand test with
partial discharge measurements (routine test)
(v) Impulse test on the neutral, if specified (special test)
(c) Test levels for method 1 are available up to 420 kV class and
for method 2 they are given up to 765 kV class. It is expected
that beyond 420 kV class, method 2 will be normally chosen.
(d) In case of switching impulse test which is intended to verify
the switching impulse voltage withstand of the line terminals
to earth and between line terminals in case of three-phase
transformers, 1.5 times rated test voltage appears between
line terminals according to prescribed test circuit. This test is
much more severe regarding higher phase-to-phase oil and
air clearances.
(e) Impulse test whether routine test or type test, is to be
conducted on all the three phases of a three-phase trans-
former. In fact, each phase is to be tested such that one phase
is in the maximum, another in the principal and the third in
the minimum voltage-tap position.
(f) Chopped wave impulse test is now a special test. Application
of shots in case of chopped wave test is more stringent now,
since two chopped waves are to be followed by two full waves,
as against one full wave as before.
(g) The rules for highest voltage winding apply to the trans-
former in toto. Transformer is to be designed and tested
according to one method alone, either by method 1 or method
2 and not by combination of the two methods for different
windings. Method of test should be known right at the enquiry
stage.
Part IV
(a) The phase marking ABC have been replaced by UVW
following the international practice.
(b) The corresponding IEC Pub 76-4 (1976) does not contain the
information concerning terminal markings and the same is
now included in IEC 60076-1.
(c) Information about neutral and connection has been included

in the connection symbols.
(d) Line terminal is now terminal No. 1 and neutral is invariably
2. In the pre-revised system, line terminal was subscripted by
2. Marking of tapping terminals now start from the line end,
instead of neutral, in the ascending order.
(e) Principal tapping, is the mean tapping position and is
essentially the full power tapping. All guaranteed parameters
usually refer to this tapping.
15.2 Other Related Standards
To supplement IS : 2026, there are a number of related standards. A

list of these is furnished in Appendix A. A few important ones are
described below:
15.2.1 IS : 1885 (Part 38)–1993 (Second Revision)

Electrotechnical Vocabulary : Power Transformer
and Reactors
This standard covers definitions of terms applicable to transformers.

The first revision of this standard published in 1977 aligns it with the
first revision of IS : 2026.
15.2.2 IS : 3639-1966 Fittings and Accessories

for Power Transformers
This standard attempts to standardize accessories and fittings for

transformers to facilitate interchangeability and rationalize use of
fittings, such as would be commonly acceptable to manufacturers
and users. Presently, it covers accessories like valves and valve
flanges, flanges for Buchholz relay, breather pipe, thermometer
pockets, level indicator, rollers and earthing terminals.
This standard is being revised to make the description elaborate
and most up to date and include other fittings like radiators, fans,
pumps, etc.
15.2.3 IS : 2099-1986 Specification for Bushings for

Alternating Voltages above 1000 V (Second Revision)
This standard covers rated values, performance requirements

and tests for bushings for three-phase alternating current systems,
having rated voltages above 1000 V and frequencies between 15 and
6000 Hz. It also provides information for ordering of bushings. How-
ever, it does not give dimensions of bushings, for which a separate
Indian Standard IS : 3347 ‘Dimensions for porcelain transformer
bushings’ which gives the dimensions of the porcelain bushings up to
123 kV ratings exists. The scope of this standard is being extended to
include condenser bushings.
The standard is based on IEC Pub 137 and British Standard BS
223.
15.2.4 IS : 8468-1977 ‘Specification for On-load Tapchangers’

This standard covers on-load tapchangers and their motor-drive
mechanisms. This is based on the first revision of IEC Pub 214 ‘on-
load tapchangers’. Besides rating and service conditions, the stan-
dard describes at length the type the routine tests to be performed
on tapchangers.
15.2.5 IS : 8478-1977 Application Guide

for On-load Tapchangers
This application guide is intended to assist in the selection of suitable
on-load tapchangers for use in conjunction with the tapped windings
of transformers.
Since the on-load tapchanger represents only a small part of the
total cost of the equipment in a power transformer, it should be
freely chosen to suit the equipment. The guiding factors for selection
of a suitable tapchanger are insulation level, current, breaking ca-
pacity, tapping positions, discharge problems, etc. The guide also
gives the information required with enquiry or order.
15.2.6 IS : 335-1993 Specificationfor New

Insulating Oils (Fourth Revision)
This standard is based on IEC 60296 (1986), which covers two classes
of oil, class I and class II. Two classes have been included in the IEC
recommendation, because of difference in natural requirements

largely dictated by climatic conditions. In this standard, however,
only one class, mostly based on class I is covered, which is considered
to be suitable for the conditions prevailing in this country.
This standard specifies the requirements for unused insulating
oil as delivered in bulk, such as tank wagons and road tankers, or
drums suitable for immersion or filling of transformers. The oil
covered by this standard are uninhibited oils free from antioxidant
additives.
The standard gives schedule of characteristics, both chemical and
electrical, which are acceptance norms for the new oil. Consequent
on the tendency for water absorption and deterioration, the limits
for power factor, resistivity and electrical strength apply to tests
made within two weeks of delivery.
If tests are not carried out within two weeks, the oil may be
treated and tested. It shall then meet the requirements.
Unused oil complying with the requirements of this standard
are considered to be compatible with one another and may be
mixed in any proportion. (Please refer to Ch. 3 “Materials Used in
Transformers”)
15.2.7 IS : 1866-2000 Code of Practice for Electrical

Maintenance and Supervision of Mineral Insulating
Oil in Equipment (Third Revision)
This standard is based on IEC 60422 (1989), ‘Maintenance and super-

vision guide for insulating oil in service and is a companion standard
to IS : 335–1993.
Insulating oil, complying with IS : 335, from the time the oil is
received, is subject to deterioration or contamination in storage, or
in handling or in service. Accordingly, a periodic treatment to main-
tain it in fit condition, is required and eventually it may have to be
replaced by new oil. This standard covers the causes of oil deteriora-
tion and recommendations are made for the various tests and their
interpretation, whether reconditioning, reclaiming replacement is
required, also giving permissible limits for satisfactory use. The
methods used for reconditioning and reclaiming are also briefly
given.
Unfortunately, no IS gives characteristics of oil to be filled in and
just after filling in the transformer, for works testing.
15.2.8 IS : 10028 Code of Practice for Selection, Installation

and Maintenance
This standard supercedes the earlier standard IS 1886–1967 which

covered information about installation and maintenance of power
and distribution transformers.
New standard IS 10028 also includes selection of transformers in
addition of installation and maintenance. It is in three parts:
Part 1–1985 Selection
Part 2–1981 Installation
Part 3–1981 Maintenance.
15.2.9 IS : 6600-1972 ‘Guide for Loading

of Oil Immersed Transformers’
This guide in the present form is applicable to oil immersed

transformers of type ONAN and ONAF. The guide indicates how oil
immersed transformers may be operated for different ambients and
duties without exceeding the acceptable limit to deterioration of
insulation through thermal effects.
Basically, the cooling medium is at 32°C (maximum weighted
average ambient) but, deviations from this are provided for in such a
way that the increased use of life when operating with a cooling
medium temperature above 32°C in summer, is balanced by the
reduced use of life, when it is below 32°C in winter.
This standard, based on IEC 60354 ‘Loading guide for oil-
immersed transformers’ is a very useful tool for the user.
15.3 New Standards
New standards being formulated are on air clearances of bushings,

application guide and testing guide on power transformer. A brief
introduction of these follows:
15.3.1 Minimum Bushing Air-clearance
IS 2026 (Part 5) 1994 titled “Transformer/Reactor Bushings-minimum

external clearances in air-specification” was published in May 1994.
This standard specifics clearances in air between live parts of

bushings of oil immersed power transformers and reactors and to
object at earth potential.
15.3.2 Applications Guide for Power Transformers
A new standard on the above subject has been published in 1983 as

IS : 10561-1983. This standard which is based on IEC. 606 (1978)
‘Application guide for power transformers’ is intended to assist in the
determination and selection of transformer characteristics. Subjects
like specification of tapping quantities, selection of winding connec-
tions for transformers for these-phase systems, parallel operation of
transformers, loading capability of the neutral point of windings in
star or zigzag connection and calculation of voltage drop (or rise) for
a specified load condition have been discussed at length.
This standard is a useful guide for purchasers at the time of
purchase and during subsequent usage.
IEC 606 is withdrawn and now available as IEC 60076 Part 8 (1997).
15.3.3 Guide to Testing of Transformers
A draft ‘Indian Standard’ is under formulation to cover allied subjects

to testing of transformers, viz. class of accuracy and frequency of
calibrations of the measuring instruments, sequence of testing,
purpose and significance of the various tests and necessary guidance
on testing, etc. This standard will be a coordinating link between
various testing codes and a useful guide for the test engineers.
15.4 Standard Specification of a Power Transformer
15.4.1 Minimum Specification of a Power Transformer
The following information should be definitely available with enquiry

and order for the design and drafting of the transformer.
(i) Type of transformer, for example separate winding trans-
former, auto-transformer or booster transformer.
(ii) Number of phases, single or polyphase.
(iii) Frequency.
(iv) Rated power (in kVA) and for tapping ranges exceeding ±5%
required power on extreme tappings.
(v) Rated voltages for each winding.
(vi) Connection symbol.
(vii) Requirement of on-load/off-circuit tapchangers or links—
number of tappings, tapping range, location of tappings,
particular voltage required to be varied and whether constant
flux/variable flux/combined voltage variation.
(viii) Impedance voltage at rated current and principal tapping for
different pairs of windings and at least on the extreme
tappings in case of parallel operation if required.
(ix) Indoor or outdoor type.
(x) Type of cooling and if different types of cooling involved, rated
power for each type of cooling.
(xi) Temperature rises and ambient temperature conditions
including altitude and in case of water cooling chemical
analysis of water.
(xii) Number of cooling banks, spare capacity if any and number of
standby cooling pumps/fans.
(xiii) Highest system voltage for each winding.
(xiv) Method of system earthing for each winding.
(xv) Whether windings shall have uniform or non-uniform
insulation, and in case of non-uniform insulation, power
frequency withstand voltage of neutral and impulse withstand
level if an impulse test on the neutral is required.
(xvi) For windings having system highest voltage greater than
300 kV, the method of dielectric testing, whether method 1 or
method 2.
(xvii) Withstand voltage values constituting insulation level of line
terminals, viz. test levels of lightning impulse, switching
impulse, one minute power frequency, long duration power
frequency with partial discharge measurement, wherever
applicable.
(xviii) Limitation of transport weight and moving dimensions and
special requirements, if any, of installation, assembly and
handling.
(xix) Whether a stabilizing winding is required.
(xx) Overfluxing conditions or any other exceptional service

conditions.
(xxi) Loading combinations in case of multi-winding transformers
and when necessary active and reactive outputs separately,
especially in case of multi-winding auto-transformers.
(xxii) Details of auxiliary supply voltage (for fans, pumps, OLTC
motor, alarm and control).
(xxiii) Controls of tapchangers.
(xxiv) Short-circuit levels of the systems.
(xxv) Vacuum and pressure withstand values of the transformer
tank.
(xxvi) Noise level requirement.
(xxvii) Number of rails and rail gauge for movement along shorter
and longer axes.
(xxviii) Fittings required with their vivid description.
(xxix) Any other appropriate information including special tests if
any and capitalization formula for the losses.
15.4.2 Need for Standardization of Specification
Transformer is a tailor-made product and requires a total effort

in the design and drafting, even if a single parameter is changed.
Standardization of the specification and design parameters of this
vital equipment of energy transport will not only help in ensuring
optimal deployment of available resources but also go a long way
in economizing the capital costs. Standard design would mean
quicker deliveries, lesser outages and a big leap towards achieving
national grid.
Efforts to standardize transformer specification have been made
in different forums. The CEA report and the CBIP specifications are
the outcome of a concerted effort over the years.
15.4.3 CEA Report of the Committee for Standardization of

the Parameters and Specifications of Major Items of
400 kV Equipment
The Government of India, in the Ministry of Energy (Department of

Power) constituted a committee with active participation of Central
Electricity Authority (CEA) in March 77 (BHEL was also
represented) to make recommendations for standardization of the

parameters and specifications of 400 kV equipment. The committee
submitted its report in April 1978.
Chapter 1 of the report gives the specification of 400 kV trans-
formers. Based on the characteristics of the available lightning
arresters, a lightning impulse withstand level of 1300 kVp and a
switching impulse withstand level of 1050 kVp have been recom-
mended for 400 kV transformers. Method 2 as per IS : 2026 (Part
III)–1981 has been adopted for specifying the insulation levels and
dielectric tests.
250 MVA rating has been standardized for the matching generator
transformers for 210 MW sets and 600 MVA generator transformer
rating is proposed for the 500 MW sets. The interconnecting 400/220
kV auto-transformers are recommended to have 250, 315, 500 and
630 MVA standard ratings. The 400/132 kV auto-transformers shall
have 100 and 200 MVA ratings. All auto-transformers have loaded
33 kV tertiary windings with one-third rating. The generator trans-
former shall have off-circuit tap switch and the auto-transformer
shall be equipped with on-load tapchanger. All other important
design parameters, viz. tapping range, impedances, operating condi-
tions, etc., and fittings and accessories have also been standardized.
Standardization is a continuous process and must go on. With this
view in mind, the reconstituted Working Group No. 1 responsible for
transformers has reviewed and updated the recommendations of
Chapter 1. This revision is much more elaborate and also included
standard dimensions of bushings from interchangeability point of
view.
15.4.4 CBIP Specifications
Central Board of Irrigation and Power (CBIP) brought out a

publication of ‘Specification for Power and Distribution Transform-
ers’ in 1968 which was revised in September 1976 to make it up to
date and more comprehensive. Chapters on testing, commissioning,
protection, etc., have also been added. It is revised again in October.
1999 and available as “Manual on Transformers, Publication No.
275.
APPENDIX I
I. Indian Standards Related to Power Transformers
(a) Specifications
1. IS : 1885 (Part: 38)–1993 Electrotechnical vocabulary, Trans-

formers and Reactors (Second revision)
2. IS : 2026 (Part I)–1977 Power transformers: Part I General (first
revision)
3. IS : 2026 (Part II)–1977 Power transformers : Part II Tempera-
ture rise (first revision)
4. IS : 2026 (Part III)–1981 Power transformers: Part III Insulation
level and dielectric tests (second revision)
5. IS : 2026 (Part IV)–1977 Power transformers: Part IV Terminal
marking, tappings and connections (first revision)
6. IS : 11171-1985 Dry type power transformers
7. IS : 2026 (Part V)–1994 Power transformers : Part V Transformer/
Reactor Bushing-minimum external clearances in air.
(b) Fittings and Accessories
1. IS : 778-1984 Gunmetal gate, globe and check valves for waterwork

purposes
2. IS : 2099-1986 Bushings for alternating voltages above 1000 V
(second revision)
3. IS : 3347 (Part I-Part VII–Section 1) Dimensions for porcelain
transformer bushings, Section 1–Porcelain parts, Voltage 3.6-123
kV
4. IS : 3347 (Part I-Part VII–Section 2) Dimensions for porcelain
transformer bushings, Section 2–Metal parts, Voltage 3.6-123 kV
5. IS : 2312-1997 Propeller type AC ventilating fans
6. IS : 3024-1965 Grain oriented electrical steel sheets and strips
7. IS : 3151-1982 Earthing transformers
8. IS : 3231-1965 Electrical relays for power systems protection
9. IS : 3401-1992 Silicagel (third revision)
10. IS : 3588-1987 Electric axial flow fans
11. IS : 3624-1987 Specification for pressure and vacuum gauges
12. IS : 3637-1966 Gas operated relays

13. IS : 3639-1966 Fittings and accessories for power transformers
14. IS : 4253 (Part II)-1980 Specification for cork composition sheets
15. IS : 6088-1988 Oil to water heat exchangers for transformers
16. IS : 7421-1988 Porcelain bushings for alternating voltage up to and
including 1000 V
17. IS : 8468-1977 On-load tapchangers
18. IS : 8603-1977 (Part I to III) Dimensions for porcelain transformer
bushings for use in heavily polluted atmosphere
19. IS : 8765-1978 Ceramic insulating materials for electrical
purposes
20. IS : 9147-1979 Cable sealing boxes for oil immersed transformers
21. IS : 9700-1991 Specification for activated alumina
22. IS : 11333-1985 Flame proof dry type transformers for use in
mines
(c) Selection, Installation, Operation and Maintenance
1. IS : 104-1979 Readymixed paint, brushing, zinc chrome, priming

2. IS : 900-1992 Code of practice for installation and maintenance of
induction motor.
3. IS : 1255-1983 Code of practice for installation and maintenance of
paper insulated power cables
4. IS : 1554-(Part I)–1988 PVC insulated heavy duty electric cables
(for working voltages up to and including 1100 V)
5. IS : 1866-2000 Code of practice for electrical maintenance and
supervision of mineral insulating oil in equipments
6. IS : 10028 Code of practice for selection, installation, maintenance
of transformers
7. IS : 2266-1989 Steel wire ropes for general engineering purposes
8. IS : 2932-1993 Enamel, synthetic, exterior undercoating and
finishing
9. IS : 3043-1987 Code of practice for earthing
10. IS : 3638-1966 Application guide for gas operated relays
11. IS : 3832-1986 Hand operated chain pulley blocks
12. IS : 3842 (Part I)–1967 Application guide for electrical relays for
AC systems. Part I, Over current relays for feeders and transformers
13. IS : 3842 (Part XII)–1976 Application guide for electrical relays for
AC systems. Part XII, Differential relays for transformers.
14. IS : 5216-1962 Guide for safety procedures and practices in

electrical work
15. IS : 5561-1970 Electric power connectors
16. IS : 13234-1992 Guide for short-circuit current calculations in 3
phase AC system
17. IS : 6034-1989 Insulating oil conditioning plant
18. IS : 6132 Shackles
Part 1–1971 General requirements
Part 2–1972 Dimensions of D-shackles
Part 3–1972 Dimensions of bow shackles
19. IS : 6600-1972 Guide for overloading of power transformers
20. IS : 7689-1989 Guide for the control of undesirable static electricity
21. IS : 8270 (Part I-V) Guide for preparation of diagrams, charts and
tables for electrotechnology
22. IS : 8478-1977 Application guide for on-load tapchangers
23. IS : 8923-1978 Warming symbol for dangerous voltages
24. IS : 9434-1992 Guide for sampling and analysis of free and
dissolved gases and oil from oil filled electrical equipment
26. IS : 9615-1980 Guide on general aspects of electromagnetic inter-
ference suppression
27. IS : 10561-1983 Application guide for power transformers
(d) Transformer Oil and Oil Testing
1. IS : 335-1993–New insulating oils (Fourth revision)

2. IS : 1448 (P : 10)–1970 Methods of test for petroleum and its
products, cloud point and power point
3. IS : 1448 (P-16) 1990 Methods of test for petroleum and its
products, cloud point and power point–Density of crude petroleum
and liquid petroleum products by hydrometer method
4. IS : 1448 (P-21)–1992 Methods of flash point by Pensky Martens
apparatus
5. IS : 1548 (P-25)-1976-Determination of kinematic and dynamic
viscosity
6. IS : 1783-1993–Drums, large fixed ends
7. IS : 1866-2000–Code of practice for Electrical, Maintenance and
Supervision of Mineral Insulating Oil in Equipment. (Third
Revision)
8. IS : 2362-1973–Determination of water by the Karl Fischer
Method
9. IS : 6103-1971 Method of test of specific resistance (Resistivity) of

electrical insulating liquids
10. IS : 6104-1971 Method of test for interfacial tension of oil against
water by ring method.
11. IS : 6262-1971 Method of test for power factor and dielectric
constant of electrical insulating liquids
12. IS : 6792-1992 Method of determination of electric strength
insulating oils
13. IS : 6855-1973 Methods of sampling for liquid dielectrics
14. IS : 9434-1992 Guide for sampling and analysis of free and
dissolved gases and oil from oil filled electrical equipment
(e) Insulation Co-ordination and High Voltage Testing
1. IS : 1876-1961 Method for voltage measurement by means of

sphere gaps
2. IS : 2071 – Methods of high voltage testing
Part 1 – 1993 General definitions and test requirements
Part 2 – 1974 Test procedures
Part 3 – 1976 Measurement devices
3. IS : 2165-1977 Insulation coordination
4. IS : 3716-1978 Application guide for insulation coordination
5. IS : 4004-1985 Application guide for nonlinear resistor type surge
arresters for AC system
6. IS : 4850-1968 Application guide for expulsion type lightning
arresters
7. IS : 6209-1982 Methods for partial discharge measurements
8. IS : 8690-1977 Application guide for measuring devices for high
voltage testing
II. IEC Publications
1. 60076-1 (2000) ‘Power transformers’

Part 1 : General
2. 60076-2 (1993) Part 2 : Temperature rise
3. 60076-3 (2000) Part 3 : Insulation levels and dielectric tests
4. 60076-5(2000) Part 5 : Ability to withstand short-circuit
5. 60214 (1989) ‘On-load tapchangers’
6. 60542 (1976) ‘Application guide for on-load tapchangers’

7. 60354 (1991) ‘Loading guide for oil immersed transformers’
8. 60551 (1995) ‘Measurement of transformer and reactor sound
levels’
9. 60296 (1986) ‘Specification for new insulating oil for transformers
and switchgear’
10. 60076-8 (1997) Application guide for power transformers.
C HAPTER 16
Loading and Life of

Transformers
D.P. Gupta
In any electric power system, right from generation to utilization,

there are several points where power transformation takes place.
The total capacity of transformers used for this purpose in a system
would be around six to eight times the installed generating capacity.
Inspite of transformers having the highest loss-efficiency among the
various electrical equipment, the cumulative energy losses are
substantial. Their construction also requires large quantities of vital
material like copper, aluminium, core and structural steels, cellu-
lose insulation, oil, etc. Considering depletion of world resources,
there is an ever increasing emphasis on conserving materials and
energy on one hand and capital on the other.
This chapter aims at bringing about a better understanding of the
behaviour of transformer under varying conditions of ambient air
temperature, load and winding temperatures and how they affect the
life of the transformer, so that it would help in conserving the vital
materials, energy and capital resources by maximizing their use
without unduly sacrificing safety, reliability and normal expected
life.
Materials can be put to optimum use by the manufacturer of the
transformer through an indepth understanding of the properties and
the stresses that are likely to occur in service. But it is the purchaser
who can optimize the utilization of the existing transformers or
funds invested in them, through maximizing their use by better un-
derstanding of the inbuilt capabilities.
It is necessary to establish a common base of understanding be-

tween user and manufacturer concerning general consequences of
loading beyond name plate rating and particular degree of loading
capability required for a specific transformer.
16.1 Life of a Transformer
A transformer has practically no moving parts, except tapchangers

or cooling fan or pump motors. Therefore, it cannot wear out like
rotating machinery. With adequate protection against corrosion, the
copper windings, laminated cores and fabricated parts will last indefi-
nitely. But insulating materials, mostly made from cellulose materi-
als, deteriorate from the effects of temperature, moisture and oxy-
gen. Out of these factors, it is the temperature which must be kept
within known limits to prevent rapid deterioration of insulation. The
time in which insulation deteriorates under normal usage to the
point of failure, may be very long by human standards. The life of a
transformer is normally dependent upon the life of the insulation.
When insulation fails, the transformer life has ended. The term
transformer life gives an impression as if it was quite definite, but in
fact a transformer hardly ever dies. It is usually killed, by some un-
usual stresses breaking down a weakened part leading to the end of
the transformer. The two factors which normally contribute to the
eventual failure are:
(a) Deterioration of insulation over a span of time with tempera-
ture, moisture and oxygen.
(b) Operating stresses, mechanical, electromagnetic, thermal,
beyond the strength of those parts which have considerably
weakened over a period of time.
The life of a transformer has ended when probability of its failure
becomes too high. It is practically impossible to determine the prob-
ability that any transformer will fail within a few months or years;
but this probability actually exists. In other words, it is not possible
to predict precisely the life of a transformer working under given
temperature and load conditions. Presently, a scientific basis for
evaluating degree of risk is not available. Dielectric strength of
insulation, immersed in oil, does not deteriorate until the material
Loading and Life of Transformers 379
has become brittle and cracked and it is, therefore, possible for a
transformer to continue to operate long after the mechanical life of
its insulation has been virtually used up, unless it is subjected to
excessive mechanical stresses like short-circuit, handling or other
mechanical shocks. It is known to users of transformers that many
times a transformer which has been giving satisfactory service for
years, when shifted to other location has failed due to damage to its
weakened insulation. It does not follow, therefore, that a trans-
former will fail when its insulation becomes brittle. However, if it is
subjected to some excessive stresses, it may fail even before the
mechanical strength is completely used up.
Insulation is deteriorating and losing its life all the time, depending
upon the temperature at which it is operating. Therefore, it becomes
necessary to understand the various factors which contribute to the
deterioration of insulation; especially the temperature and laws
relating to ageing of insulating materials, so that a transformer can
be put to optimum use without too rapid a loss of life, resulting in
premature failure.
16.2 Ageing of Insulation
In oil immersed transformers, paper or cellulose material along with

oil forms the major insulation. Therefore, the insulation must main-
tain adequate dielectric strength against voltage surges and ad-
equate mechanical strength against short-circuit forces.
Paper and pressboard insulation when heated under oil for long
periods of time, lose mechanical strength but dielectric strength is
hardly affected until the paper is charred to the point where free
carbon becomes conducting or too brittle to withstand mechanical
shocks. Deterioration to this extent is complicated by free water
liberated by decomposition, leading to lowering of dielectric strength.
Depolymerization of insulation takes place when deterioration
sets in.
Long before this occurs, mechanical strength will be reduced
considerably. There are several ways of defining mechanical
strength of paper, but the most meaningful and easily measured
quantity is tensile strength. Although there is no simple relationship
between loss of tensile strength and loss of effective transformer life,
it seems reasonably logical to use loss of tensile strength as a

measure of life.
16.2.1 Causes of Loss of Tensile Strength of Paper in Oil
The combination of factors to which paper insulation is subjected to

in service are several, including temperature, time and presence of
moisture, oxygen and various other reaction products in oil. Many
studies have been made in the past which give reasonable closeness
of relationship between tensile strength, temperature and time.
However, it has been found that very small amounts of moisture
accelerate the loss of tensile strength (Fig. 16.1). Presence of oxygen
with moisture can lead to various kinds of reactions depending upon
temperature and presence of catalytic materials.
16 X103
15
14
13
12
11
10
Tensile strength
6
5
2 0.
27 0.1 0.055
1% % wa
1.1
%
2.2%
1 ter
%
0
200 400 600 800 1000 2000 3000
Hours of ageing at 120°C
Figure 16.1 Effect of moisture on mechanical strength of oil-

immersed paper at 120°C in sealed container.
16.3 Law of Insulation Ageing
The insulation of a transformer tends to age and deteriorate when

heated. The higher the temperature, the faster is the insulation
deterioration. During periods of subnormal operating temperature,
the loss of life of the insulation will be less than normal. But when
the operating temperatures are greater than normal, the loss of life
will be higher than normal. Consequently, transformer may be
safely operated for a time at above normal temperatures, provided
the loss of insulation life during this period is adequately compen-
sated for by operation for a sufficiently long time at temperatures
lower than normal.
The law of insulation ageing is not linear. It is now generally
accepted that the life of insulation deteriorates with temperature
and is governed by the law of Arrhenius:
Life = e(A + B/T) (1)
where A and B are constants, (derived by experiment for a given
insulating material), T is the absolute temperature and e is the
Napierian base. In the range of 80° to 140°C winding hot-spot, this
law can be expressed in a more convenient form called Montsinger
relation:
Life = e–pq (2)
where p is a constant and q is the temperature in °C. Various
investigators have not always agreed on the length of life at any
given temperature. However, they do agree, that between 80° to
140°C, the rate of loss of life due to ageing of transformer insulation
is doubled for every 6°C rise in temperature. Till few years back,
the insulation ageing was based on 8°C rule, which has been replaced
by 6°C rule, considered more near the actual behaviour of cellulose
insulation under varying temperatures.
The Montsinger relation can be used to obtain the relative rate of
using life at any hot-spot temperature over the normal hot-spot
temperature (98°C) at rated lood and reference ambient and tem-
perature rise conditions dealt in Chapter 5, paragraphs 5.1 and 5.2.
Relative rate of using life,
Rate of using life at any hot-spot q c
V=
Rate of using life at rated hot-spot q cr
= 2(qc – qcr)/6
= 2(qc – 98)/6
or qc = 98 + 19.93 log10 V (3)
If a transformer operating at hot-spot of qc uses up one day’s life
(based on concept of normal life at 98°C hot spot) in t hours:
t = 24/V
= 24 ¥ 10(98–qc)/19.93 (4)
Values of q c for various values of t in hours are given in Table 16.1:
Table 16.1 Values of qc for Various Values of t
t, hr/day qc
24 98
16 101.5
12 104
8 107.5
6 110
4 113.5
3 116
2 119.5
1.5 122
1.0 125.5
0.75 128
0.5 131.5
16.3.1 Concept of Weighted Ambient

As the law of insulation ageing is not linear (Fig. 16.2), an equal
variation of ambient over the mean in a given period of time and
constant load will result in corresponding variations in the winding
hot-spot temperatures and a net loss of insulation life over the nor-
mal. Therefore, the temperature which, if maintained continuously
during the period of time under consideration, would result in the
same ageing of insulation as that occurring under the actual condi-
tion of ambient temperature variations, is called weighted ambient
temperature. The weighted ambient temperature is, therefore, not
the arithmetic mean value of the ambient temperatures over a given
period of time, and has essentially the element of life built in it.
8
Relative rate
q°C of using life
5
80 0.125
86 0.25
92 0.50
98 1.00
4
104 2.0
110 4.0
116 8.0
3 122 16.0
Rate of insulation ageing

128 32.0
134 64.0
140 128.0
2
Loading and Life of Transformers
(Normal) 1
0
80 86 92 98 104 110 116
Winding hot-spot temperature, °C
383
Figure 16.2 6°C Law of insulation ageing.

16.4 Significance of Weighted Value

of Ambient Temperature
16.4.1 Ambient Temperature

Any location on the surface of the earth sees variations in the ambi-
ent temperatures over the day, month and full year due to the rota-
tion of earth on its own axis in 24 h and around the sun in a year. If
the temperature variations over a day are plotted for most of the
parts of the world, the variation observed would be found close to
sinusoidal (Fig. 16.3). Similarly, the variation of daily average ambi-
ent temperature plotted over the whole one year is near sinusoidal
(Fig. 16.4). It will be seen that if the daily variation is superimposed
on the yearly variation, it will appear as 365 sinusoidal ripples on the
yearly sinusoidal variation curve (Fig. 16.5). The monthly variation
which is only one-twelfth part of the yearly ambient variation curve,
could be treated as a straight line variation for all practical purposes.
16.4.2 Availability of Ambient Temperature Data

for a Location
The data on ambients for a given location can normally be obtained
from the records maintained by the Meteorological Departments for
the past several years located in kolkata, Delhi, Madras, Nagpur and
°C
1 1
New Delhi:– 20 to 21
82 82
20
Source: IMD thermograph
No. 100/65
15 Daily mean 13.7 °C
10
5
1 1
20 21
82 82
0 10 12 14 16 18 20 22 24 2 4 6 8 Hrs
Figure 16.3 Typical daily ambient variation.

°C
New Delhi
40
Source: IMD clomato-
logical table
30
Mean of mean monthly 25.2 °C
(Annual mean)
20
10
J F M A M J JY A S O N D Months
(Jan. to Dec.)
Figure 16.4 Typical mean ambient variation over a year.
Daily variation
Max. ambient variation
Mean ambient variation
Min. ambient variation

Ambient temp.
Annual mean
A M J JY A S o N D J F M Months
(April to March)
Figure 16.5 Pattern of ambient variation over one whole year.
Poona. Information on the extreme maximum and minimum ambi-

ent over a number of years, mean of daily maximum and
minimum over a month, highest and lowest in a month is available
for several locations in British Meteorological office publication
Tables of relevant humidity and precipitation for the world, and also
in the Indian Meteorologically Department Climatological Tables of

Observatories for 228 locations based on 30 years’ records from 1931
to 1960.
In the following paragraphs, the use and interpretation of the
available data is given towards developing ambient, load and life re-
lationship.
16.5 Relationship Between Weighted Ambient,

Winding Rise and Hot-spot Temperatures
For evaluation of optimum loading with life of transformer as basis,

it is necessary to understand the relationship between ambient,
winding rise and winding hot-spot.
For ready reference, ambient temperatures as specified in
IEC : 76-1 (1976) and IS : 2026 (Part I)–1977 are given below:
Table 16.2 Reference Ambients
IS : 2026 IEC–76
1. Max. ambient air temperature 50°C 40°C
2. Max. daily average air temperature 40°C 30°C
3. Max. yearly weighted air temperature 32°C 20°C
4. Max. cooling water temperature 30°C 25°C
5. Average cooling water temperature in a day 25°C —
Also, for ready reference, temperature rise limits for oil immersed
transformers with class A insulation as specified in IEC 76-2 (1976)
and IS : 2026 (Part II)–1977 are:
Table 16.3 Reference Temperature Rises
I. Air Cooled IS : 2026 IEC 76

1. Winding (a) ONAN
rise ONAF
ONAF (Non-directed) 55°C 65°C
(b) ODAF (Directed) 60°C 70°C
2. Top oil (a) When sealed or 50°C 60°C
with conservator
(b) When not so sealed or 45°C 55°C
without conservator
Contd.
Table 16.3 Contd.
II. Water Cooled IS: 2026 IEC 76

1. Winding (a) OFWF (Non-directed) 60°C 65°C
rise (b) ODWF (Directed) 65°C 70°C
2. Top oil rise (a) When sealed or with 55°C 60°C
conservator
(b) When not so sealed or 50°C 55°C
without conservator
Basically, the cooling medium temperature as laid down by IS :

2026 (Part I)–1977 is 32°C and by IEC 76–1 (1976) is 20°C, over which
deviations are provided for, in such a way that the increased use of
life when operating with cooling medium temperature above these
values (as in summer) is balanced by the reduced use of life when
operating below these values (as in winter).
Experience indicates that normal life is some tens of years. It can-
not be stated precisely, because it may vary even between identical
units, owing in particular to operating factors which may differ from
one unit to another.
16.5.1 Ambient–Winding Hot-Spot Relationship

A simple thermal diagram of relationship between ambient, winding
and oil rises and winding hot-spot is shown in Fig. 16.6. The actual
distribution of temperature is complex, but the difference is not con-
sidered sufficiently significant to invalidate the method. The follow-
ing assumptions are made:
(a) The oil temperature increases linearly up the winding.
(b) The average oil temperature rise is the same for all the
windings of the same leg.
(c) The difference of temperature between the oil at the top of
the winding (assumed equal to that of the top oil) and the oil
at the bottom of the winding (assumed equal to that at the
bottom of the cooler) is the same for all windings.
(d) The average temperature rise of the copper at any position up
the winding, increases linearly parallel to the oil temperature
rise, with a constant difference Dqwo between the two straight
lines: see Fig. 16.6 (Dqwo being the difference between the
average temperature rise by resistance and the average oil
temperature rise).
(e) The average temperature rise of the top portion of the

winding will be the sum of top oil temperature and Dqwo.
(f) The hot-spot temperature rise is higher than the average
temperature rise of the top portion of the winding. To take
account of the difference between these two temperature
rises, a value of 0.1 Dqwo is assumed for natural oil circulation.
[Forced oil circulation has been dealt with in IEC Publication
354 (1972) Appendix A, Sub-clause 1.2.2.] Thus the
temperature rise of the hot-spot is equal to the top oil
temperature rise plus 1.1 Dqwo.
The ambient-winding hot-spots in IS and IEC specifications on
transformers are related to the reference temperature rises as fol-
lows:
I. For Natural Oil Circulation ONAN/ONAF
Table 16.4
Description IS : 2026 IEC 76

(a) Max. yearly weighted temp. 32°C 20°C
(b) Top oil rise *45/50°C *55°C/60°C
(c) Average winding rise by resistance 55°C 65°C
(non-directed)
(d) Mean oil rise ONAN/ONAF assumed 36/40°C 44/48°C
0.8 ¥ Top oil rise
(e) Winding gradient Dqwo (c – d) 19/15°C 21/17°C
(f) Hot-spot gradient 10% of Dqwo 1.9/1.5°C 2.1/1.7°C
(g) Hot-spot temperature at rated load 97.9/98.5°C 98.1/98.7°C
(a + b + e + f)
(h) Hot-spot temperature for unity life 98°C 98°C
* For non-conservator or unsealed transformers.
II. For Forced Oil Circulation OF AF, OFWF

Generally, in forced oil cooled transformers the difference between
the inlet the outlet oil temperature rise will be less than with
natural circulation. Therefore, with winding temperature rise of
65°C (IEC) and 55°C (IS) the hot-spot temperature will be lower than
98°C. However, for all practical purposes, 98°C can be considered as
applicable for forced cooled (non-directed) transformers also.
However, as it may be more economical to use higher current

density and lower oil rise, for the purpose of framing IEC loading
guide, top oil temperature rise of 40°C and a hot-spot temperature
rise of 78°C at rated load has been assumed as most onerous. Yet,
another difference in thermal time constant of such transformers
which is assumed as 2 hr and not 3 hr as for natural cooled trans-
formers, would make the loading guide tables different. The IS :
6600–1972 Guide for loading oil immersed transformers does not
cover the forced oil cooled transformers at present.
The patterns of variation of temperatures in a transformer with
constant load and varying temperatures and also with varying loads
and constant ambient have been shown in Fig. 16.7(a) and 16.7(b). In
fact, in practice the variation will be a combination of both varying
load and varying ambient (Fig. 16.7(b)). A simplified load diagram for
cyclic duty is given in Fig. 16.8(b).
16.5.2 Rated Load and Weighted Ambient

The transformer name plate rating is based on the load that can
be carried continuously without effecting the normal life expectancy
of a transformer at annual weighted ambient air temperature,
provided the temperature of the ambient air and the oil and winding
rises are within the specified limits of IS : 2026/IEC 76 as applicable
Winding
0.8 X
Temperature
Oil
t
Ambient
Load
6 12 18 24
Time Hrs
Figure 16.7(a) Steady load v. varying ambient.

Temperature
ing
ind
W
Oil
ient
Amb
Oil
Ambient
Load
6 12 18 24
Time
Figure 16.7(b) Varying load v. varying ambient.
Av. WDG
rise (top)
Dqwo Hot-spot
Top oil
Top of WDG
0.1Dqwo
Average oil Dqwo

Av. WDG
rise
Bottom oil
Bottom of WDG
Temp. rise
Dqwo–Diff. between AV. WDG rise & AV. oil rise
Figure 16.8(a) Constant ambient and varying load.
to a particular case. The winding hot-spot temperature at the refer-

ence ambient temperatures will be 98°C, which will rise and fall with
the variation of ambient air temperature, but to a lesser extent,
depending upon the thermal time constant of the transformer and
with a certain time lag. Even though the law of insulation ageing is
not linear, for any variation of hot-spot temperature above and below
Load as fraction of rated power
K2
K1 K1
Time, Hrs 24 Hrs
Figure 1.6.8(b) Simplified load diagram for cyclic duty.
98°C hot-spot, the expected life of transformer will remain unaffected

over the normal as the hot-spot is based on the weighted ambient
temperature, which by definition takes into account the different
rates of ageing at the elevated and subnormal temperatures. In
other words, if the transformer loading, whether daily, monthly or in
any other convenient combination of time periods, is adopted on the
basis of corresponding weighted ambient temperatures applicable
over the period, the actual ambient variation would be of no
consequence and can, therefore, be ignored. At the same time, it is
necessary that whatever loading scheme; whether daily, monthly or
in any other combination is followed, it should be carried throughout
the entire one year, as the reference annual weighted ambient
temperature is based on one full cycle of ambient temperature varia-
tion, resulting from one complete rotation of the earth round the
sun. It is also important not to confuse the weighted ambient with
the arithmetic mean ambient temperature. The difference between
the two has been dealt with in the following paragraphs.
16.5.3 Difference between Arithmetic Mean

and Weighted Mean Temperature
The essential difference between the two is that, whereas the
weighted ambient is tied to the laws of insulation ageing, the arith-
metic mean ambient is not. A close look at the curve in Fig. 16.2 will
show that for an equal variation of hot-spot temperature over a given

period on either side of 98°C hot-spot, will result in a net loss of
insulation life over the normal, as the rate of ageing during excess
temperature is more rapid than with the temperature below 98°C. In
other words, the resultant rate of loss of life over the entire period
will be more than the normal rate of loss of life at 98°C. To illustrate
this point, consider temperature plot in Fig. 16.9. The ambient varia-
tion in a day is such that it results in a transformer hot-spot of 104°C
for 12 hr and 92°C for the rest of 12 hr. Based on the relationship
between the hot-spot temperature and the rate of loss of life as
shown in Fig. 16.2, the average rate of loss of life in this case over a
period of 24 hr will be
12 ¥ 2 + 12 ¥ 0.5
= 1.25
24
units. The corresponding value of hot-spot temperature for 1.25 rate
of loss of life from Fig. 16.2 will be 100°C and not the arithmetic
mean temperature of 98°C. In other words, for the same loss of life,
the value of weighted ambient temperature, will be 100°C–98°C = 2°C
higher than the arithmetic mean value in this case.
Hot-spot temp.
104
Hot-spot temperature, °C
100 Weighted temperature

98 Airthmetic mean
temperature
Hot-spot temp.
92
86
0 12 24
Time, Hrs
Figure 16.9 Weighted ambient and arithmetic mean ambient.

16.5.4 Correction Factors and Range of Temperature Variation

(a) Correction Factor
This difference between the weighted ambient and the arithmetic
mean ambient has been termed as correction factor. Because of the
nonlinearity of the law of ageing, the correction factors will be differ-
ent for different shapes of ambient variation curves.
(b) Range of Temperature Variation

It will be observed from the plots of the daily and yearly ambient
temperatures, that the pattern of variation over a day and a year is
near sinusoidal (Figs 16.3 and 16.4) and over the month it can be
approximated to a straight line (Fig. 16.10). The difference between
the extremes of the ambient variation has been termed as the range
of ambient variation.
(c) Correction Factors for Sinusoidal

and Straight Line Variation of Ambients
The correction factors for sinusoidal and straight line variation of
ambients which are necessary to calculate the weighted ambient
qext qmax.
Extreme Mean of daily max
maximum in a month over several years
over several years Envelopes of daily
maximas, for several years
Plot of daily variations

qA
Max. daily variation
Monthly Max. variation in

weighted daily mean in a month
Temperature
daily mean
Envelopes of daily
minimas, for several years
qmean
qmin
Mean of mean max. & min.
Mean of daily min. (monthly mean)
over several years
1 week 2 week 3 week
Month
Figure 16.10 Ambient variation over a month for several years.

have been worked out mathematically in Annexure I. The correction

factors for various ranges of ambient variations; sinusoidal or
straight line, pattern have been given in Fig. 16.11 (Curve A for
straight line variation and curve B for sinusoidal variation).
12
11
10
9
B
Correction factor °C
8 A
n
7 a tio
ari on
6 lv iati
o ida ar
us linev
5 Sin ht
aig
4 Str
3
2
1
0
10 20 30 40 50
Temperature range °C
Figure 16.11 Correction factor.
(d) Correction Factors for Stepped Variation of Weighted Ambients

To arrive at the value of the weighted ambient applicable to the
entire period covered by the two groups of months with two different
weighted ambients (Annexure I), an appropriate correction from
Fig. 16.12 is to be subtracted from the higher of the two weighted
ambients. Following this approach further will give the value of
annual weighted ambient for the place.
16.6 Determination of Weighted Ambient Temperature
16.6.1 Annual Weighted Ambient Air Temperature in India

No ready reference or documented information on this subject is
available except one paper by Gupta and Awasthi. The practical diffi-
culties arise in working out annual weighted ambient, because it is
6 k = correction
y = range K = 0.5
5 Weighted
Correction ‘c’ in °C
ambient
4 h1
h2
K = h1/h2 K = 1.0
3 H
2 K = 2.0
5 10 15 20 C°
Range ‘d’ in °C
Figure 16.12 Correction factor for stepped variation of weighted ambient.
extremely difficult to predict the weather conditions from one year

to the next with any degree of accuracy. The safest procedure, there-
fore, is to assume that the weighted ambient temperature adopted
will be the highest, which has ever occurred, and if higher tempera-
tures do occur, they will be only slightly higher. The weighted ambi-
ent temperatures can be calculated from the data on weather over
several years available in the meteorological offices. For India, the
annual mean temperature has been worked out as 26.7°C (say 27°C),
by Shri H.R. Kulkarni in his paper on Standardization of ambient
temperatures. The published records of temperatures in India also
indicate that the maximum variation in the daily mean ambient tem-
perature over a year is not likely to exceed 23°C and the maximum
variation of ambient air temperature in a day by more than 23°C.
Therefore, applying correction factor from curves in Fig. 16.11, the
conservative correction on account of these two variations, will
be about 5°C, assuming the two variations as sinusoidal. Actually,
the daily variation in ambient temperature affects transformer
temperature by about 80% due to the thermal time constant of the
transformer.1
Therefore, the correction for daily variation can be applied from
the curve A for the straight line variation (Fig. 16.11) which is 80% of
the curve B values for the sinusoidal variation. The maximum annual
weighted ambient for India, therefore, can thus be safely fixed at
27°C + 5°C = 32°C. [IS : 2026 (Part I)–1977 and IS : 6600–1972].
16.6.2 Monthly Weighted Air Ambient Temperature

for a Given Location
It is essential to know the weighted air ambient temperature for a
given location for the entire one year or parts thereof, normally
months, over which optimum loading of a transformer could be de-
termined. Considering that most of the ambient records available
with the Meteorological Department, are either monthly or yearly,
in this para a method has been suggested for determining the
weighted air ambient temperature employing the available data.
The weighted ambient air temperature for a month is based on
the following three important ambient values (Ref. Fig. 16.10):
(a) Monthly mean value of the ambient
(b) Maximum variation in the daily mean ambient over a month
(c) Maximum ambient variation in any day of the month
It will be reasonably safe to consider the worst of the three values
over a number of years, which can be obtained from the IMD
records. The most conservative value of the weighted ambient for the
month will be the sum of the monthly mean ambient and the appro-
priate correction factors for the two variations in ambient referred to
above. Whereas correction for variation in the daily mean ambient
temperature over a month which could be approximated to a straight
line variation will be applied from curve A of Fig. 16.11, the correc-
tion for sinusoidal variation of ambient temperature over a day will
also be applied from curve A and not curve B of Fig. 16.11. This is
based on investigations, which have shown that the actual variations
in the transformer temperatures are about 80% of the total variation
in the ambient air temperature in a day, due to the thermal time
constant of the transformer. As the correction factors in curve A are
about 80% of those from curve B, correction for daily variation could
be safely applied from curve A. It will be observed from Fig. 16.10
that for a month, daily variation touches maximum and minimum
every day. The mean of daily maxima and minima for the month
occurring over a number of years are given in the IMD tables, from
which monthly mean ambient temperature can be worked out. But
utilizing the data on mean of highest temperatures in the given
month, it is possible to get an approximate idea of (a) the maximum
variation in the daily mean over one month and (b) the maximum
daily variation over one month. Correction factors for the variations
(Fig. 16.11) when added to the mean monthly ambient will give
a reasonably safe monthly weighted ambient temperatures. (Ref.
Annexure II for worked out examples).
16.6.3 Weighted Ambient for a Group of Months in a Year

To obtain the value of weighted ambient for a group of months hav-
ing two different weighted ambients, a further correction is neces-
sary. The mathematical derivation is given in Annexure I. Curves in
Fig. 16.12 give correction factors. The correction, read from the ap-
propriate curve is to be deducted from the higher of the two weighted
ambients to arrive at the resultant weighted ambient, applicable to
the total period of time of the two load groups.
However, it is important that whatever scheme of loading for a
transformer is adopted, it is followed for the entire period for which
the weighted ambient is applicable and not a part thereof.
It will be seen from the Annexure III that the weighted ambient
temperatures are considerably lower in winter months than in the
summer. It will, therefore, be advantageous to load a transformer on
the basis of monthly weighted ambient temperatures where the
heavy peak loads are met in winter, otherwise to cater for summer
peak loads it would be more advantageous to load the transformers
on annual weighted ambient basis. This point has been explained in
the worked out tables for New Delhi at Annexure II (Sheet 3) and
also Fig. 16.13.
16.6.4 Ambient Temperature to be Considered in

Tables of Loading Guides. Determination
of Effective or Weighted Value of Cooling
Medium Ambient Temperature
If q a varies appreciably during the high loading time t, then a
weighted value of qa should be used, because the weighted ambient
temperature will be higher than the arithmetic average.
Consider operation at constant load with a varying ambient tem-
perature qa for a given period t. The weighted ambient temperature
q¢a during that period is given by the formula
1
1
2q¢a/6 =
t z
0
2q a / 6 dt
If the time t is divided into N equal intervals, the preceding formula

becomes
N
1
2q ¢a/6 =
N Â 2q a /6
1
% of Sr
(rated New Delhi
load)
120
110 Monthly permissible

load
Annual permissible
100 load
90
J F M A M J JY A S O N D
Months (Jan. to Dec.)
qa
New Delhi
40
30 Annual weighted
ambient
20 Weighted ambient
monthly
10
0
J F M A M J JY A S O N D
Months (Jan. to Dec.)
Figure 16.13
From this, the weighted ambient temperature

F1 2 IN
q a /6
q ¢a = 6 log 2 GH N Â JK
1
which is practically equal to
LM 1 10
N
q a / 20
OP
= 20 log10
MN N Â
1 PQ (5)
For a transformer operating on constant load throughout the year, a

weighted ambient temperature can be fixed once for all in a given
climate. This must be such that the faster deterioration during sum-
mer is exactly balanced by the slower deterioration in winter. In a
temperature climate it is found that the annual weighted mean is
generally 5°C higher than the arithmetic mean.
Example 16.1 Using monthly average values (more accurately us-

ing monthly weighted values) for qa:
LM 1 10
12
q a/ 20
OP
Annual weighted ambient temperature = 20 log10
MN 12 Â
1 PQ
qa = 30°C for 2 months
20°C for 4 months Average 15°C Weighted
10°C for 4 months average 19.8°C
0°C for 2 months
16.7 Relationship between Weighted Ambient and Load
Once the weighted ambient has been determined for a given loca-
tion, the corresponding loading capabilities of a transformer can be
worked out from the following relationship between weighted ambi-
ent, oil and winding rises and load.
qc = qa + Dqb0 + (Dqb – Dqb0) (1 – e–t/t ) + (Dqcr – Dqbr)k2y
x
D qb = Dqbr
F 1 + dk I
2
GH 1 + d JK
where
qa = Weighted ambient temperature (cooling medium)
qc = Winding hot-spot temperature at K load = qa + Dqc
qcr = Winding hot-spot temperature at rated load
Dqc = Winding hot-spot temperature rise at K load
Dqcr = Winding hot-spot temperature rise at rated load
qbo = Original oil temperature
qb = Top oil temperature at K load = qa + Dqb
qbr = Top oil temperature at rated load
Dqb = Top oil temperature rise at load K
Dqbr = Top oil temperature rise at rated power
t= Time duration in hours of any load power
t= Oil-air thermal time constant of transformer at rated load,
in hours
S
K = Ratio of any load/rated load =
Sr
S1
K1 = Ratio of initial load/rated load =
Sr
S2
K2 = Ratio of permissible load/rated load =
Sr
x = Exponent for oil rise
y = Exponent for winding rise
d = Ratio of load loss at rated load/no-load loss (Assumed 5 for
the guides)
a = Subscript representing ambient (cooling medium)
b = Subscript representing top oil
c = Subscript representing hot-spot of winding
r = Subscript representing rated value
From the equation, the permissible load can be worked out by
substituting the values for the various parameters. For convenience,
a load versus weighted ambient curve has been plotted (Fig. 16.14)
which applies to ONAN/ONAF/OFAF cooling with nondirected flow
in transformers with or without conservator or sealing. Following
similar lines of approach, other weighted ambient versus load curves
can be drawn for different types of transformer coolings as above
relationship is universal and applicable to any type of cooling.
16.8 Alternative Approach for the Calculation

of Weighted Ambient
16.8.1 Monthly and Annual Weighted Ambients

The approach given in Sec. 16.7 above requires considerable data
and lengthy calculations. A simpler, sufficiently conservative and
practical approach could be adopted by adding fixed correction of 5°C
(which allows for maximum variation in the mean and maximum
variations in daily temperature over a month or year 23°C each) to
the maximum monthly mean ambient temperature to obtain the
monthly weighted ambient temperature. The value of maximum
mean ambient temperature for a month, can be obtained from the
IMD records by taking a mean of the mean maximum and minimum
daily temperature for each month. A 5°C correction is safe enough
as brought out by the worked out examples in this chapter. In fact
150
140
130
Note: IS: 6600-1972

is presently
120 for ONAN/ONAF
110 IS:
660
0
IEC
Permissible load as % of rated load

100 35
4
IEC ONAN/ONAF
354
90
OFAF/OFWF
32°C ONAN/ONAF
0°C 10°C 20°C 30°C 40°C
Weighted ambient
Figure 16.14 Weighted ambient load relationship.

401
correction of 5°C will be more conservative for places near the sea,
where temperature variations are less than the plains. Annexure III
is based on the above approach and could be used for optimizing the
use of transformers in the 48 locations. Similar loading tables can be
worked out for another 180 locations in India from the available tem-
perature records in the IMD climatological tables, of observatories in
India (1931-60).
16.8.2 Tables for Permissible Loads on a Transformer

A few worked out examples have been given in Annexure II, for de-
termining the monthly and annual weighted ambients and loads fol-
lowing the more detailed approach dealt with in Sec. 16.7.
Permissible loads on a transformer to IS : 2026 with ONAN/ONAF
cooling, have been worked out in detail for 48 locations in India by
following alternative approach for calculating the weighted ambients
given in Sec. 16.8. The 48 locations have been chosen from the IMD
climatological tables, which provide ambient data for about 230
locations, so as to cover most of the states in India. In the choice of
locations, coastal areas, extremely hot desert areas and some hill
stations have been included to bring out their characteristic pattern.
The loading tables which are sufficiently safe and conservative, pro-
vide values of monthly and annual weighted temperature and corre-
sponding loads for the 48 locations on the assumption, that the trans-
formers with ONAN/ONAF cooling, strictly comply with IS : 2026.
Some of the interesting observations which emerge from the An-
nexure III permissible loads are as follows:
(a) Out of 48 locations the permissible annual load for 43
locations is within ±2% of the rated capacity of transformers.
(b) Where peak-loads are heavier in winter, than in summer
months, it will be found advantageous to follow permissible
monthly loads; but where summer loads are heavier than
winter months, use of permissible annual load may be more
advantageous.
(c) For determining permissible daily loads on a transformer,
IS : 6600 Guide for loading for oil immersed transformers can
be followed using table for appropriate monthly weighted
ambient.
(d) Coastal locations do not show any appreciable difference in
monthly permissible loads over the year compared to the
locations in the interior.
(e) The maximum annual weighted ambient is 34°C for Madurai

and minimum is 18.4°C for Srinagar.
(f) The maximum monthly weighted ambient is 41.2°C in May
for Ramagundam and Kota, followed closely by Jhansi,
Nagpur and Bikaner. The value of minimum monthly
weighted ambient tem-perature is 6°C in January for
Srinagar.
(g) For Kolkata, Cuttack, Khammam, Madras, Madurai,
Ramagundam, Surat and Tiruchy, the weighted monthly
ambient is more than 32°C for eight months, the longest
period in a year.
(h) The highest extreme ambient temperature of 49.4°C occurred
in May in Bikaner, followed by 48.3°C in May/June in Agra,
Hissar, Lucknow and Ludhiana.
(i) The permissible loads given in the Annexure III sheet 1 and
2 are safe loads for transformers, based on weighted ambients
and indicated as a percentage of the rate value. As the
correction factor of 5°C to arrive at the monthly weighted
ambient is over-applied by 2°C, the permissible monthly
loading can be increased by another 2% over the figures given
in the table, if such need arises in actual use.
16.8.3 Temperature Indicator Setting

under Overload Conditions
Most of the transformers are provided with oil temperature indica-
tors (OTI) for measuring hot top oil temperature of a transformer.
Winding temperature indicators (WTI) are also used to give an idea
of the winding hot-spot temperature by thermal image method.
Under steady rated load conditions and ambients assumed at ref-
erence level (IS/IEC), the top oil temperature will be between 77°C
and 82°C depending upon whether transformer is without or with
conservator, and winding hot-spot temperature of 98°C. In practice,
however, neither the loads are steady nor the ambients. According
to loading guides which are based on normal cyclic duty, load current
should not exceed 1.5 times the rated value. Further, overloads shall
not result in hot-spot temperature exceeding 140°C and top oil tem-
perature 115°C. (The limit of 115°C for the temperature has been set
bearing in mind that the oil may overflow at oil temperatures above
normal).
As already discussed in the earlier parts of this chapter, any varia-

tion in the daily air ambient temperature results in about 80% varia-
tion in the transformer temperature. This will be considered while
working out ultimate top oil and hot-spot temperatures likely to oc-
cur during the period under consideration. For loading conditions as
laid down in IS and IEC loading guides, it may be safer to set the OTI
and WTI alarm, 100°C OTI and 115°C WTI and trip OTI 105°C and
WTI 125°C. This will allow a sufficiently safe margin for any extreme
cooling medium temperature condition, which may occur once in a
while over the limits, on the basis of which the weighted ambient
was calculated. For setting of OTI and WTIs, manufacturers’ recom-
mendations could be followed. And if need arises, they could be al-
tered by calculating the limits under any other load and ambient
conditions.
16.8.4 Effect of Altitude

The transformer standards lay down the limits for the operation of
transformers at different altitudes because of the variation of these
limits with the density of the air. It is to be noted that in the prepara-
tion of tables at Annexure III, no allowance has been made for alti-
tude for locations like Shillong, Simla, Srinagar, Kathmandu and
other hill stations. Therefore, before using these tables a suitable
allowance/correction will have to be made for altitude.
16.8.5 Limitations Imposed by Other Considerations

The permissible loading of a transformer as discussed in this chapter
and in the guides, is restricted to thermal considerations. Other
economic factors depending upon capitalization of losses, etc. will
have to be considered separately. The transformer that has been
operating at loads higher than the rated may not comply with the
thermal requirements on specified short-circuit limits in IEC or IS.
Before allowing over-loads it will also be necessary to check the rat-
ings of the terminal outlets, tapchange devices and similar other
attachments.
The loading possibility of various fittings should be determined by
referring to the manufacturer and if there are constraints, lower
limit of loading and duration will have to be adopted. It would also be
necessary to ascertain the thermal capability of the associated equip-
ment like cables, bus-bars, circuit breakers, current transformers,
etc. It has also to be ensured that the winding hot-spot temperature
does not exceed 140°C owing to accelerated deterioration effects,

either because the formation of deterioration products is too fast for
them to be taken away by the oil or because a gaseous phase is
started, sufficiently rapid to lead to over-saturation and the forma-
tion of bubbles which may endanger the electric strength. Limitation
of conservator size or air cells, may also restrict the overload. The
present guides are applicable to transformers up to 100 MVA ratings.
However, work is being done by an IEEE task force to develop a
loading guide for transformers above 100 MVA ratings, for Working
Group on Guides for Loading Insulation Life Sub-committee of IEEE
Transformer Committee.
In recent times a strong interest has been shown in the functional
life characteristics of large transformers loaded above name
plate ratings. Cost would be prohibitive for performing life test on
transformers above 100 MVA ratings, but some research projects
have been started towards gaining an insight into factors affecting
loadability. The principal factors influencing the approach adopted by
IEEE Task Force are:
(a) Stray flux heating was identified to be a possible limitation to
transformers rated above 33 MVA (single-phase) and 100 MVA
(three-phase) when operated in excess of name plate rating.
Stray flux heating, which is function of both the MVA rating
and the leakage reactance of the transformer, is caused by
leakage flux which finds a path outside the iron core. The heat
generation typically occurs in the tank and structural parts
of the transformer, but can also occur in the core and the
winding leads.
(b) Research work performed in the areas of the combined effects
of thermal ageing and short-circuit stresses on dielectric
strength leads to a conclusion that life expectancy could be
reduced in larger transformers with high short-circuit forces.
(c) The phenomenon of bubble evolution from cellulose insula-
tion at overload temperatures is presently an area of major
research. The results of research to date indicate that
bubbling does significantly reduce both power frequency and
impulse strength of transformer insulation and is recognized
as an important limitation when operating above name plate
rating.
(d) Because the large transformers usually have higher voltage
ratings, they tend to have greater average dielectric stresses
and larger volumes of insulation under stress. Also, they are

often specified with reduced steps of BIL. These factors,
together inherently reduce the margins for dielectric
deterioration.
(e) There is a wide variation in user’s requirements for loading
beyond nameplate rating for the large transformers. In
addition, there appears to be a trend to use planned loading
above nameplate rating as contrasted to the more conser-
vative use of emergency loading.
(f) Many misunderstandings and problems have arisen from
reference to the present loading guide in a transformer
specification to identify loading capability requirements. The
present guide does not convey the necessary information
required by a transformer designer, to provide the above
nameplate rating capabilities in the transformer that are
desired by the user.
(g) The consequences of failure of these large transformers
are becoming a matter of more concern with the rapidly
increasing outage and repair cost. Hence, it is of great
importance that the most complete and up-to-date infor-
mation should be available for planning loads above
nameplate ratings.
16.8.6 Effect of Loading beyond Nameplate Rating

When a transformer is designed for operation at a given rating level,
stress factors are selected to achieve a satisfactory life expectancy at
that rating level. If the rating level is exceeded, there is some en-
croachment on margins and the question of risk of premature failure
must be considered. While the quantification of the risk factor may
be difficult, as a first step it is necessary to identify the potential
effects from increased loading. Some of the important areas of con-
cern during transformer loading beyond nameplate rating are:
(a) Evolution of gas from overheated insulation on winding and
lead conductors which may jeopardize dielectric integrity.
This factor is a relatively new concern which has been
addressed in the recent literature.
(b) Evolution of gas from insulation adjacent to metallic parts
heated by stray electromagnetic flux.
(c) Deterioration of the electrical and mechanical properties of

conductor insulation from thermal ageing. This is the
traditional life characteristic on which calculations are
furnished in existing guides.
(d) Cumulative thermal degradation of mechanical properties of
structural insulation parts, plus some temporary reduction of
mechanical capability at elevated temperatures.
(e) Permanent deformations of conductors, insulation materials
or structural parts produced by thermal expansion at overload
temperatures.
(f) Gasket rupture in bushings produced by internal pressure
buildup at current levels above rated capability.
(g) Degrading contact resistance in tapchangers resulting from
buildup of oil decomposition products at a localized contact
point hot-spot. This could only be a concern with prolonged
operation beyond nameplate rating.
(h) Ancillary equipment internal to the transformer which may
be subjected to effects similar to those mentioned above.
(i) Oil spills produced by oil expansion during operation at
temperatures above the design value. This is an inconvenience
rather than an operating risk.
16.9 Transformer Loading Guides
Loading guides for power transformers date back to 1942 with very
little change in basic philosophy. In the currently published guides,
the emphasis is still on a long-time end of life failure mode associated
with the thermal deterioration of mechanical and dielectric proper-
ties of solid insulation in the windings. The useful life of a trans-
former has been arbitrarily equated to the loss of 50% of the initial
tensile strength of winding conductor insulation. Consumption of life
may be calculated according to equations or tables given in the
guides based on Arrhenius type relationship between the degrada-
tion rate and temperature. Although there are other stresses like
mechanical, dielectric and thermal produced by stray fluxes, which
increase with MVA rating of transformers, consideration of these
stress factors has generated an agreement in the industry, that a
reasonable point of separation for the loading guide would be 100
MVA. The present guides, therefore, are limited. In the available
transformer loading guides the basic approach is the same, except

that the loading tables have been modified to suit the local conditions
in a country. Some of the salient factors of relevance to this chapter
based on the study of IEC 354 (1972) and IS : 6600-1972 have been
discussed in the paragraphs below:
(a) The guides are applicable to transformers up to 100 MVA,
which complying with the relevant IEC 76 or IS : 2026 for
transformers are operated within the specified limits. Any
deviation must be considered carefully and such changes
made in loading as become necessary for the safety of the
transformers and other associated equipment.
(b) For transformers above 100 MVA, advice of manufacturer
should be followed. (IEEE task force is working on the guide
to loading of oil immersed transformers above 100 MVA and
also IEC Pub. 354 is being revised to include transformers
above 100 MVA).
(c) Under no circumstances, hot-spot temperature be allowed to
exceed 140°C beyond which gaseous phase starts and may
endanger the electric strength.
(d) Top oil temperature should be kept within 115°C and matched
to suit the expansion space for oil.
(e) To use the guide, knowledge of weighted ambient tempera-
tures; daily, monthly or annual for a given location, is
necessary. Practically, these values can easily to worked out
from the meteorological records for various months and the
year as discussed in this chapter. As the loading table in the
guide are based on annual weighted ambient temperature,
the daily loads on the transformer can be met by referring to
the appropriate table in the guide, which corresponds to the
annual weighted ambient, where transformer is actually
located.
(f) Similarly, daily loads can be met by referring to the tables in
guide on the basis of monthly weighted ambient. Loading on
the basis of monthly weighted ambient then will have to be
followed for the entire one year and not changed to annual
weighted in the middle, as integration of load-loss relation-
ship is over full one year.
(g) The transformer rated ONAN and ONAF can be loaded as per
appropriate set of tables in the guide, in terms of rated power
for ONAN, if fans are not working and rated power for ONAF
if fans are working. Similarly, transformers with both ONAN
and OFAF cooling can be dealt with.
(h) Tables in the loading guides give permissible cyclic daily duty
with a normal life consumption of 24 hr per day, equal to that
consumed during 24 hr at 98°C hot-spot. If actual load
diagram has two or more periods of high load separated by
periods of low load, the high-load time can be taken as the
sum of all high-load times. This condition of intermittent high
loads is less onerous than single overload for the same total
time.
(i) For a given load cycle, and given ambient conditions, tables
can be used to determine the rating of the transformer
needed for the load.
16.10 Loading by Hot-Spot Temperature Measured by WTI
With the knowledge of relative rate of use of life at any hot-spot tem-
perature as discussed in Sec. 16.3 of this chapter, for practical pur-
poses, consumption of life on any day could be worked out for a trans-
former for any loading and ambient condition by integrating the con-
sumption of life based on the record of readings of the winding tem-
perature indicator at regular intervals over the day. The winding
temperature indicator indicates thermal image of the winding hot-
spot temperature and can be considered sufficiently accurate when it
is corrected to the actual measured values based on temperature
rise test on the transformer. In adopting this approach, care should
be taken to ensure:
(a) Oil temperature and winding temperature at any point of
time do not exceed the specified limits of 115°C and 140°C
respectively.
(b) Upper limit of load as dictated by the loading limits of other
associated components shall not be exceeded. In any case,
loads above 150% of the rated are special emergency
conditions.
(c) Process of life integration when done manually should be over
a day, in such a way that resultant consumption of life is
related to one day’s life at 98°C hot-spot (Normal day).
(d) It would be safer to adopt the highest value of hot-spot

reached in any time interval for manual integration of life.
(e) The above method of life integration by using WTI readings
takes care of any changes in ambient and load and effect of
thermal mass.
Example 16.2 Calculation for working out the life of a transformer

consumed on a day, on the basis of hourly WTI readings is given in
the table below:
qc = Hot-spot Rate of Life used

Hours (Hourly WTI using life in hours
reading °C)
1 80 0.125 0.125
2 78 0.100 0.10
3 79 0.11 0.11
4 80 0.125 0.125
5 81 0.13 0.13
6 86 0.25 0.25
7 93 0.55 0.55
8 98 1.00 1.00
9 104 2.00 2.00
10 104 2.00 2.00
11 101 1.40 1.40
12 93 0.55 0.55
13 98 1.00 1.00
14 103 1.70 1.70
15 104 2.00 2.00
16 104 2.00 2.00
17 98 1.00 1.00
18 93 0.55 0.55
19 98 1.00 1.00
20 101 1.40 1.40
21 101 1.40 1.40
22 95 1.00 1.00
23 93 0.55 0.55
24 83 0.20 0.20
Total 22.69 Hr
From the above table it will be seen that against the normal al-
lowed consumption of life of 24 hr actual consumption has been 22.69
hr, which leaves a balance of 1.31 hr of unused life.
16.11 Selection and Use of a Transformer
In the guides to the loading, the load diagrams have been simplified
into low-load and high-load periods. Based on the value of weighted
ambient qa, appropriate loading tables in the guide can be used.
However, if qa lies between two tables, either select the nearest one
above or interpolate between the closest two tables.
For transformers rated for both ONAN and ONAF cooling, for
ONAN cooling use appropriate tables in terms of the rated power for
ONAN cooling, and in terms of the rated power of ONAF cooling, if
the fans are brought into operation. Similarly, transformers with
ONAN and OFAF cooling will be dealt with or without oil pumps in
operation.
It is recommended that pumps and fans are put into service before
the overloading occurs, in order to have the winding hot-spot tem-
perature low enough to slow down the ageing process. The power
taken by these auxiliaries is at least partially compensated for by the
decrease in load loss, resulting from a low temperature. Following
worked out examples will help towards a better understanding of the
concept of life and loading:
Example 16.3 Determine maximum permissible load which a 15

MVA, ONAN cooled transformer to IS : 2026 can take for 6 hr, if
the initial load on it was 12 MVA. The weighted ambient tempera-
ture is 20°C.
Cooling type : ONAN
qa : 20°C
K1 : 0.8 (12 MVA of rated 15 MVA)
*t : 6 hr
*(Designated h in IS : 6600)
K2 (from Table 3 of IS : 6600) = 1.25
\ Maximum permissible = 1.25 ¥ 15 = 18.75 MVA load for 6 hr.
The overload will be reduced to 12 MVA for the remaining
part of the day.
Example 16.4 Same transformer as in Example 16.3, but with

initial load of 7.5 MVA. Determine the maximum permissible load
for 2 hr.
Cooling type : ONAN
qa : 20°C
K1 : 0.5 (7.5 MVA of rated 15 MVA)
t : 2 hr
\ K2 (from Table 3 of IS : 6600) = 1.55
But as the upper limit of load is restricted to 1.5, the maximum
permissible load the transformer could be allowed to carry for
2 hr = 1.5 ¥ 15
= 22.5 MVA
before load is reduced to 7.5 MVA for the remaining part of the day.
Example 16.5 Find out the rating of ONAN cooled transformer to

IS : 2026, which will cater for a lean load of 8 MVA for 18 hr and 14
MVA for balance 6 hr in a day in a location with weighted annual
ambient temperature of 32°C.
K2 14
= = 1.75
K1 8
A life with K2/K1 = 1.75 can be plotted intersecting the curves in
Table 5 of IS : 6600. Where the line intersects the 6 hr curve, the
K2 reads = 1.14.
S2 14
Sr = (Normal transformer rating) = = = 12.28 MVA.
K2 114
.
Example 16.6 What will be the normal rating of ONAN/ONAF

cooled transformer to IS : 2026, which will deliver continuously a
load of 12.7 MVA at a location with annual weighted temperature
of 30°C.
Cooling type : ONAN/ONAF
qa : 30°C
t : 24
K2 : 1.02 (From Table 4 of IS : 6600)
S2
K2 =
Sr
S2 12.8
Sr = = = 12.5 MVA
Kr 1.02
Example 16.7 A 20 MVA ONAN/ONAF transformer to IS : 2026 is

located in New Delhi, where peak loads are heavier in winter than
in summer. Find out the safe permissible loads it can carry in
each month.
Method 1
The permissible load curve vis-a-vis weighted ambient is given in
Fig. 16.14 (also refer Annexure II, Sheet 2). This is based on addition
of a flat correction of 5°C on the mean monthly ambient to obtain
monthly weighted ambient and corresponding permissible load from
curve in Fig. 16.14.
Method 2
The monthly weighted ambients can be determined by adding cor-
rections worked out on the basis of IMD records (Ref. Annexure II,
Sheet 3) and permissible load from curve in Fig. 16.14. This method
permits slightly higher overloads than method 1.
REFERENCES
1. Equivalent Ambient Temperature for Loading Transformers by W.C.
Sealey of Allis Chalmers (USA).
2. Transformer Engineering, A Treatise On Theory, Operation and
Application of Transformers by GE (USA).
3. Transformers for Electric Power Industry by Westinghouse (USA).
4. IS : 2026 : Specification for Power Transformers.
5. IS : 6600 : Guide for Loading of Power Transformers.
6. Kulkarni H.R., Paper on Standardization of Ambient Temperatures.
7. IEC Loading Guide for Oil Immersed Transformers, IEC publication
354.
8. Gupta D.P., and L.C. Awasthy, Weighted Ambient Temperature as the
Basis for Loading Transformer, read in CBIP Research Session at
Jaipur.
9. Progress Report on a Guide for Loading Power Transformers Rated in
excess of 100 MVA, IEEE Paper Vol. PAS 100, 8 Aug. 1981.
ANNEXURE I
Correction Factors
Sheet No. 1
A. 1.1 Correction Factor for Sinusoidal Variation
If ambient temperature varies in a sine wave (Fig. A1.1), with duration of

cycle as H, the weighted ambient (correction over base temperature) denoted
by x can be worked out from the following relationship:
H
1
2x/6 =
H z0
2 q / 6 ◊ dh (i)
where q = qm sin h
H = 2p
Therefore, Eq. (i) can be re-written as
2p
1
2x/6 =
2p z
0
2q m
sin h / 6
dh (ii)
The expansion of 2qm sin h is :
q m sin h q sin 2 h
2qm sin h/6 = 1 + loge 2 + m (loge 2)2 + ....
6 62 – 2
x
Base
H
A
qm
Figure A1.1
Neglecting higher powers of qm and solving Eq. (ii), value of x is given by

x = 19.93 log10 (1 + 0.0033q 2m) (iii)
For convenience, results have been plotted as curve B of Fig. 16.11.
Sheet No. 2
A.1.2 Correction Factor for Straight Line Variation
A similar approach as above can be followed for working out a correction for
straight line variation also.
In the case of straight line variation as shown in Fig. A.1.2, at any instant
the value of q can be written as
2q m
q = h
H
(value of h at any instant must bear the correct sign). If x is the weighted
ambient temperature (correction over base temperature), it can be repre-
sented by the following relationship:
1 +H / 2 q /6
2x/6 =
H z-H / 2
2 ◊ dh
2q m h
(i)
1 +H /2
or 2x/6 =
H z
-H /2
2 6H
◊ dh (ii)
2q m
Function h can be expanded and written as follows by ignoring
6H
higher powers of qm :
2q m 4q 2m (log e 2) 2 2
22qm/6H ¥ h = 1 + (loge2) h + h + ....
6H 36 H 2 2
Substituting value of 22q mh/6H in Eq. (ii) and solving for x:
x = 19.93 log10(1 + 0.0022 q 2m) (iii)
for convenience of working, results have been plotted as curve A of Fig. 16.11.
+qm
–H/2
+H/2
–qm
Figure A.1.2
A.1.3 Correction Factor for Stepped Variation
If two weighted ambient temperatures qm and qm – y are maintained for a

period h1 and h2 respectively, and x denotes the factor which should be
subtracted from qm to get the weighted ambient applicable to total time H,
the relationship between various parameters is
Sheet No. 3
2(qm – x)/6 =
1 RS z h1
2qm / 6 dh + z H
2 (qm - y)/6 dh
UV (i)
H T 0 h1 W
Assuming h1/h2 = K and rewriting Eq. (i)
1
2(qm – x)/6 = {2qm/6 kh2 + 2(qm – y)/6h2}
(1 + k)h2
2qm / 6
= {k + 2–y/6} (ii)
1+ k
y
Expanding 2y/6 = 1 + log 2 + ... = 1 + 0.115 y
6
Rewrite Eq. (ii) as
RS
2–x/6 = 1 -
1 UV
T (1 + k)(1 + 8.7 / y) W
x = – 19.93 log 1 -
RS 1 UV (iii)
T (1 + k)(1 + 8.7 / y) W
For convenience, working results have been plotted as curves in Fig. 16.12.
Weighted ambient
qm
h1 h2
H
Figure A.1.3
ANNEXURE II
Sheet No. 1
Location - Agra Annexure-II
Source - IMD Table Page 115 Sheet No-1
Sr.
No. Description J F M A M J JY A S O N D Annual
A. Mean of daily max. 22.7 25.7 31.9 37.7 41.8 40.5 34.8 32.8 33.2 33.3 29.2 24.1 32.3°C
B. Mean of daily min. 7.4 10.3 15.7 21.6 27.2 29.5 27.0 25.8 24.6 19.1 12.0 8.2 19.0°C
C. Mean monthly 15.1 18.0 23.8 29.6 34.5 35.0 30.9 29.3 28.9 26.2 20.6 16.2 25.7°C
D. Max of variation in 2.9 5.8 5.8 4.8 0.5 4.1 1.6 0.4 2.7 5.6 4.4 1.4
Mean in the month
E. Mean of highest in 27.2 31.5 37.8 42.5 45.4 45.0 40.7 35.9 36.5 36.4 33.0 28.1
the month over
several years
F. Max. daily variation 21.4 21.2 22.2 21.0 21.4 16.0 18.0 12.8 12.4 14.8 20.4 22.4
2 ¥ (E – C – D/2)
G. Correction for ‘D’ 0.1 0.2 0.2 0.2 – 0.1 – – – 0.2 0.2 –
H. Correction for ‘F’ 1.9 1.9 2.1 1.8 1.9 1.1 1.4 0.6 0.6 0.9 1.7 2.1
80% of curve
I. Weighted ambient
C+G+H= 17.1 20.1 26.1 31.6 36.4 36.2 32.3 29.9 29.5 27.3 22.5 18.3
G + H assumed 5°C = (20.1) (23.0) (28.8) (34.6) (39.5) (40.0) (35.9) (34.3) (33.9) (31.2) (25.6) (21.2)
J. Correction for ~ (Curve B) = 2.6°C

max. variation in
mean (monthly)
U|
over one whole
year.
V|
C max - C min 35 – 15.1 = 19.9°C W
K. Correction for
max. daily
U| (Curve A) = 21°C
variation in one
whole year
V|
F max. 22.4°C W Total correction = 4.7°C
L. Annual weighted
C+J+K= 25.7 + 4.7 = 30.4°C
Daily (30.7°C)
Values in (c)
F max
by adding 5°C
D J correction
A E Mean monthly
CB Monthly Yearly
Plot 30 days Plot 12 months
Sheet No. 2
Location : Bombay Annexure II
Source : IMD table, Page 211 Sheet No. 2
Ser.
A. Mean of daily max. 29.1 29.5 31.0 32.3 33.3 31.9 29.8 29.5 30.1 31.9 32.3 30.9 31.0
B. Mean of daily min. 19.0 20.3 22.7 25.1 26.9 26.3 25.1 24.8 24.7 24.6 22.8 20.8 23.6
C. Mean of monthly 24.2 24.9 26.8 28.7 30.1 29.1 27.4 27.2 27.4 28.2 27.6 25.8 27.3
D. Max. variation in 0.7 1.9 1.9 1.4 1.0 1.7 0.2 0.2 0.8 0.6 1.8 1.6
Mean in the month
E. Mean of highest 33.0 33.9 35.0 34.8 34.6 34.4 31.8 31.1 31.7 35.1 34.8 33.7 36.7
in the month over
several years
F. Max daily variation 16.8 16.0 14.4 10.8 8.0 9.0 9.0 7.6 7.8 13.2 12.6 14.2
2 ¥ (E – C – D/2)
G. Correction from ‘D’ – – – – – – – – – – – –
H. Correction for ‘F’ 1.2 1.1 0.9 0.4 0.2 0.3 0.3 0.2 0.2 0.7 0.6 0.8
80% of curve
I. Weighted ambient 25.4 26.0 27.7 29.1 30.3 29.4 27.7 27.4 27.6 28.9 28.2 26.6
C+G+H= (29.2) (29.9) (31.8) (33.7) (35.1) (34.1) (32.4) (32.2) (32.4) (33.2) (32.6) (30.8)
J. Correction for
max. variation in
mean (monthly)
over one whole
year
C max – C min. 30.1 – 24.2 = 5.9°C CF = 0.2°C
:
K. Correction for
max. daily
variation over
one whole year
F max. 16.8°C
CF = 1.2°C
Total corr. 1.4°C
L. Annual weighted
C+J+K=
Daily 28.7°C
(32.3°C)
F max
D J
A E Mean monthly
CB Monthly Yearly
Sheet No. 3
Location – New Delhi Annexure - II
Source :- IMD table page 127 Sheet No-3
Sr.
B. Mean of daily min 7.3 10.1 15.1 21.0 26.6 28.7 27.2 26.1 24.6 18.7 11.8 8.0 18.8
C. Mean monthly 14.3 16.8 22.6 28.6 33.5 34.3 31.2 29.9 29.4 25.9 20.2 15.7 25.2
mean in the month
E. Mean of highest in 27.0 29.6 35.8 41.2 44.3 44.1 40.3 37.2 37.0 36.3 32.4 26.0 44.9
the month over
several years
F. Max. daily variation 23.0 19.8 20.4 20.2 20.8 16.4 17.0 14.2 11.7 15.2 20.0 19.2
2 ¥ (E – C – D/2)
G. Correction for D – 0.1 0.1 0.1 – 0.1 – – 0.1 0.1 0.1 –
H. Correction for F 2.2 1.7 1.8 1.8 1.9 1.1 1.3 0.8 0.5 1.0 1.7 1.6
80% of curve
I. Weighted ambient 16.5 18.6 24.5 30.5 35.4 35.5 32.5 30.7 30.0 27.0 22.0 17.3
C+G+H=
J. Correction for
max. variation in
mean (monthly)
over one whole
year.
C max – C min: 34.3 – 14.3 = 20°C CF = 2.6°C
:
K. Correction for
max. daily
variation over 23.0°C CF = 2.2°C
:
one whole year
F. max Total correction = 4.8
L. Annual weighted 30.0

C + J + K = 30.2 Daily 30.2
F max
D J
A E Mean monthly
CB Monthly Yearly
Sheet No. 4
Location : Srinagar Annexure II
Source IMD table page 131 Sheet No– 4
Sr.
B. Mean of daily min. –2.3 – 0.8 3.5 7.4 11.2 14.4 18.4 17.9 12.7 5.7 – 0.1 – 1.8 7.2
C. Mean monthly 1.0 3.6 8.4 13.4 17.9 21.7 24.6 23.9 20.5 14.2 7.7 3.5 13.4
mean in the month
E. Mean of highest 9.4 13.7 20.1 26.3 31.5 34.4 35.5 34.0 32.6 28.5 20.4 14.4 36.0
in the month over
several years
F. Max. daily
variation
2 ¥ (E – C – D/2) 14.2 15.4 18.4 21.4 22.8 21.6 18.8 19.4 20.8 22.2 19.0 17.6
G. Correction for D – 0.1 0.1 0.1 0.1 0.1 – – 0.1 0.2 0.2 0.1
H. Correction for F 0.8 1.0 1.5 1.9 2.1 2.0 1.6 1.7 1.8 2.0 1.6 1.3
80% of curve
I. Weighted Ambient
C+G+H= 1.8 4.7 10.0 15.4 20.1 23.8 26.2 25.6 22.4 16.4 9.5 4.9
J. Correction for
max variation
in mean (monthly)
over one whole
year
C max – C min 24.6 – 1.0 = 23.6 CF = 3.3°C
:
K. Correction for
max daily
variation over
CF = 2.1°C
one whole year 22.8°C
F. Max 5.4
L. Annual weighted Daily

C+J+K= 18.8 °C
F max
D J
A E Mean monthly
CB Monthly Yearly
ANNEXURE III
Loading Table Transformers ONAN/ONAF Cooled (To IS : 2026)
Sheet No. 1
Location Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Annual
Agra 111 108 102 97 92 92 96 98 98 101 105 110 101

Ahmedabad 106 103 100 96 94 94 98 99 98 99 102 105 100
Ajmer 111 108 103 98 96 94 98 99 99 102 106 110 102
Allahabad 110 107 102 96 92 93 97 98 98 100 105 109 101
Ambala 112 110 104 99 94 93 96 97 97 102 107 111 102
Amritsar 114 111 107 102 97 94 96 97 98 102 108 113 102
Bangalore 105 102 101 100 100 102 103 103 103 103 104 105 102
Baroda 106 102 100 96 94 95 98 99 99 100 102 105 100
Belgaum 104 102 100 99 100 102 103 103 103 102 103 104 102
Bhopal 108 105 102 98 93 96 100 101 101 102 105 108 102
Bikaner 112 109 103 98 92 91 94 97 96 100 106 111 101
Bombay 102 102 100 98 97 98 100 100 100 99 99 101 100
Calcutta 106 103 99 96 96 97 98 98 98 99 102 105 100
Coimbatore 102 101 99 98 98 100 101 101 101 101 102 102 101
Cuttack 104 102 98 95 94 96 98 98 98 99 102 104 99
Cochin 100 100 98 98 99 100 101 101 101 100 100 100 100
Dehradun 113 111 107 102 98 98 100 101 102 104 109 112 104
Dhanbad 108 105 101 96 95 96 99 99 99 101 104 107 101
Gauhati 108 106 103 101 100 99 98 98 98 101 104 107 102
Hissar 112 109 104 99 94 92 95 96 97 101 106 111 102
Hyderabad 104 102 100 97 95 98 101 101 101 102 104 105 101
Jabalpur 108 106 102 98 93 95 100 100 100 102 106 108 102
Jaipur 111 108 103 98 94 94 97 99 99 101 105 109 102
Jammu 113 110 105 100 95 93 96 98 98 102 106 111 102
421
Notes: Safe loads for transformers based on Weighted Ambient (Figures in %).
Sheet No. 2
422
Location Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Annual
Jhansi 109 106 102 96 91 92 97 99 99 101 105 109 101

Kathmandu 116 114 109 106 103 102 102 102 103 106 111 115 107
Khammam 102 100 97 95 92 95 98 99 98 99 102 103 98
Kota 108 105 100 95 91 92 97 98 98 99 103 107 100
Karnool 102 100 97 94 93 97 99 99 99 100 102 104 100
Lucknow 110 107 102 97 93 93 97 98 98 101 105 109 101
Ludhiana 113 110 105 100 94 93 96 97 97 101 107 111 102
Madras 102 101 99 97 94 95 96 97 97 99 101 102 98
Madurai 101 100 98 96 95 96 96 97 97 98 100 101 98
Marmugao 101 102 100 98 98 100 101 101 101 100 100 101 100
Mysore 104 102 100 99 100 102 103 102 102 102 103 104 102
Nagpur 105 102 99 95 91 94 99 100 100 99 104 106 100
New Delhi 112 109 103 98 94 93 96 97 98 101 106 110 102
Patna 109 106 101 97 95 95 97 98 98 100 103 108 101
Transformers, 2/e
Poona 105 103 101 98 97 100 102 102 102 101 103 105 102
Puri 104 102 100 98 97 97 98 98 98 99 102 104 100
Ramagundam 102 101 97 93 91 94 99 99 99 99 102 103 98
Shillong 116 114 110 107 106 105 105 105 106 109 113 116 109
Simla 120 118 115 111 107 106 108 108 109 112 115 117 112
Srinagar 124 122 117 113 108 104 102 102 105 112 117 122 113
Surat 103 102 99 96 96 97 99 99 99 98 100 102 99
Tiruchy 102 100 98 96 95 95 95 97 97 99 101 102 98
Trivandrum 100 100 99 98 99 100 101 101 100 100 100 100 100
Vizag 103 102 100 98 96 96 98 98 98 99 102 103 100
Notes: Safe loads for transformers based on Weighted Ambient (Figures in %).
Sheet No. 3
S. No. Location Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Annual
qext 31.1 35.6 42.8 45.0 47.2 48.3 45.6 42.2 40.6 41.1 36.1 30.0 48.3
qmax 22.7 25.7 31.9 37.7 41.8 40.5 34.8 32.8 33.2 33.3 29.2 24.1 32.3
qmin 7.4 10.3 15.7 21.6 27.2 29.5 27.0 25.8 24.6 19.1 12.0 8.2 19.0
1. Agra qmean 15.1 18.0 23.8 29.6 34.5 35.0 30.9 29.3 28.9 26.2 20.6 16.2 25.7
qa 20.1 23.0 28.8 34.6 39.5 40.0 35.9 34.3 33.9 31.2 25.6 21.2 30.7
% Rated
Load 111 108 102 97 92 92 96 98 98 101 105 110 101
qext . 36.1 40.6 43.9 46.2 47.8 47.2 42.2 38.9 41.7 42.8 38.9 35.6 47.8
qmax 28.7 31.0 35.7 39.7 40.7 38.0 33.2 31.8 33.1 35.6 33.0 29.6 34.2
qmin 11.9 14.5 18.6 23.0 26.3 27.4 25.7 24.6 24.2 21.2 16.1 12.6 20.5
2. Ahmedabad qmean 20.3 22.75 27.2 31.35 33.5 32.7 29.4 28.2 28.6 28.4 24.6 21.1 27.4
qa 25.3 27.8 32.2 36.4 38.5 37.7 34.4 33.2 33.6 33.4 29.6 26.1 32.4
% Rated
Load 106 103 100 96 94 94 98 99 98 99 102 105 100
qext 31.7 35.6 41.7 44.6 45.6 45.6 44.4 40.0 40.6 38.9 35.0 31.1 45.6
qmax 22.2 25.3 30.7 35.9 39.5 38.1 33.2 30.9 32.1 32.9 28.9 24.4 31.2
qmin 7.3 9.9 15.7 21.9 27.3 27.7 25.6 24.3 23.7 17.8 10.9 7.7 18.3
3. Ajmer qmean 14.8 17.6 23.2 28.9 30.7 32.9 29.4 27.6 27.9 25.4 19.9 16.0 24.8
qa 19.8 22.6 28.2 33.9 35.7 37.9 34.4 32.6 32.9 30.4 24.9 21.0 29.8
% Rated
Load 111 108 103 98 96 94 98 99 99 102 106 110 102
423
Sheet No. 4
424
qext 31.1 36.1 41.7 45.0 47.2 47.8 45.6 40.0 39.4 40.6 35.6 31.3 47.8
q max 23.7 26.7 33.3 38.8 42.1 39.8 33.6 32.6 32.8 32.6 29.0 24.8 32.4
qmin 9.1 11.6 17.0 22.5 27.4 28.9 26.6 26.0 25.2 20.4 13.1 9.3 19.8
4. Allahabad qmean 16.4 19.2 25.2 30.6 34.8 34.4 30.1 29.3 29.0 26.5 21.0 17.0 26.1
qa 21.4 24.2 30.2 35.6 39.8 39.4 35.1 34.3 34.0 31.5 26.0 22.0 31.1
% Rated
Load 110 107 102 96 92 93 97 98 98 100 105 109 101
qext 28.9 33.9 41.7 45.0 47.8 47.8 46.7 43.9 40.6 39.4 35.6 29.4 47.8
q max 20.8 23.8 29.6 36.2 40.8 40.5 35.2 33.8 35.4 33.2 28.6 23.2 31.8
qmin 6.8 8.5 14.1 19.7 24.9 27.3 26.0 25.4 23.9 16.4 10.2 7.1 17.5
5. Ambala qmean 13.8 16.2 21.8 28.0 32.8 33.9 30.6 29.6 29.6 24.8 19.4 15.2 24.6
qa 18.8 21.2 26.8 33.0 37.8 38.9 35.6 34.6 34.6 29.8 24.4 20.2 29.6
% Rated
Load 112 110 104 99 94 93 96 97 97 102 107 111 102
Transformers, 2/e
qext 25.0 32.2 35.6 43.3 46.1 46.7 45.6 40.0 40.6 38.3 32.2 27.7 46.7
q max 18.6 22.6 27.5 34.2 38.9 40.4 35.6 34.2 34.4 31.9 26.5 24.4 30.5
qmin 4.5 6.5 11.5 16.2 21.4 25.2 25.9 25.3 23.8 16.6 8.8 5.0 15.9
6. Amritsar qmean 11.6 14.6 19.5 25.2 30.2 32.8 30.8 29.8 28.8 24.2 17.6 13.2 23.2
qa 16.6 19.6 24.5 30.2 35.2 37.8 35.8 34.8 33.8 29.2 22.6 18.2 28.2
% Rated
Load 114 111 107 102 97 94 96 97 98 102 108 113 102
Sheet No. 5
qext 32.2 34.4 37.2 38.3 38.9 37.8 33.3 33.3 33.3 32.2 31.1 31.1 38.9
q max 26.9 29.7 32.3 33.4 32.7 28.9 27.2 27.3 27.6 27.5 26.3 25.7 28.8
qmin 15.0 16.5 19.0 21.2 21.1 19.7 19.2 19.2 18.9 18.9 17.2 15.3 18.4
7. Bangalore qmean 21.0 23.1 25.6 27.3 26.9 24.3 33.2 23.2 23.2 23.2 21.8 20.5 23.6
qa 26.0 28.1 30.6 32.3 31.9 29.3 28.2 28.2 28.2 28.2 26.8 25.5 28.6
% Rated
Load 105 102 101 100 100 102 103 103 103 103 104 105 102
qext 35.6 41.7 43.3 45.9 46.7 45.6 40.0 37.2 41.1 41.7 38.3 36.1 46.7
q max 30.1 32.4 36.6 39.9 40.7 37.2 32.4 31.5 32.6 35.0 33.4 31.0 34.4
qmin 10.8 12.7 16.6 21.7 26.1 27.1 25.4 24.8 24.1 19.9 14.3 11.4 19.6
8. Baroda qmean 20.4 22.6 26.6 30.8 33.4 32.2 28.9 28.2 28.4 27.4 23.8 21.2 27.0
qa 25.4 27.6 31.6 35.8 38.4 37.2 33.9 33.2 33.4 32.4 28.8 26.2 32.0
% Rated
Load 106 102 100 96 94 95 98 99 99 100 102 105 100
qext 33.3 37.2 39.4 39.5 40.6 37.2 32.0 31.7 33.9 33.6 32.8 34.6 40.6
q max 30.1 32.2 35.0 35.7 34.0 27.5 25.2 25.6 27.0 30.1 29.3 29.3 30.1
9. Belgaum qmin 14.0 15.1 18.0 19.5 20.6 20.6 19.8 19.4 19.0 18.6 17.1 13.9 18.0
qmean 22.0 23.6 26.5 27.6 27.3 24.0 22.5 22.5 23.0 24.4 23.2 21.6 24.0
qa 27.0 28.6 31.5 32.6 32.3 29.0 27.5 27.5 28.0 29.4 28.2 26.6 29.0
% Rated
Load 104 102 100 99 100 102 103 103 103 102 103 104 102
425
Sheet No. 6
426
qext 32.2 36.1 40.0 44.2 45.6 43.9 40.6 35.0 36.1 37.8 33.3 32.8 45.6
q max 25.7 28.5 33.6 37.8 40.7 36.9 29.9 28.6 30.1 31.3 28.5 26.1 31.5
qmin 10.4 12.5 17.1 21.2 26.4 25.4 23.2 22.5 21.9 18.0 13.3 10.6 18.5
10. Bhopal qmean 18.0 21.5 25.4 29.5 33.6 31.2 26.6 25.6 26.0 24.6 20.9 18.4 25.0
qa 23.0 26.5 30.4 34.5 38.6 36.2 31.6 30.6 31.0 29.6 25.9 23.4 30.0
% Rated
Load 108 105 102 98 93 96 100 101 101 102 105 108 102
qext 31.1 37.2 42.8 47.2 49.4 48.9 47.2 43.3 43.9 42.2 37.2 32.2 49.4
q max 22.3 26.1 31.8 37.6 42.0 41.7 38.9 36.1 36.6 35.3 30.3 24.7 33.6
qmin 5.0 8.2 14.4 20.8 27.2 29.3 28.0 26.6 25.2 18.6 10.0 5.6 18.2
11. Bikaner qmean 13.6 17.2 23.1 29.2 34.6 35.5 33.5 31.4 30.9 27.0 20.2 15.2 25.9
qa 18.6 22.2 28.1 34.2 39.6 40.5 38.5 35.4 35.9 32.0 25.2 20.2 30.9
% Rated
Load 112 109 103 98 92 91 94 97 96 100 106 111 210
Transformers, 2/e
qext 35.0 38.3 39.7 40.6 36.2 37.2 35.6 32.3 35.0 36.6 36.2 35.1 40.6
q max 29.1 29.5 31.0 32.3 33.3 31.9 29.8 29.5 30.1 31.9 32.3 30.9 31.0
qmin 19.4 20.3 22.7 25.1 26.9 26.3 25.1 24.8 24.7 24.6 22.8 20.8 23.6
12. Bombay qmean 24.2 24.9 26.8 28.7 30.1 29.1 27.4 27.2 27.4 28.2 27.6 25.8 27.3
qa 29.2 29.9 31.8 33.7 35.1 34.1 32.4 32.2 32.4 33.2 32.6 30.8 32.3
% Rated
Load 102 102 100 98 97 98 100 100 100 99 99 101 100
Sheet No. 7
qext 31.9 36.7 41.1 43.3 43.7 43.9 36.7 36.1 36.1 35.6 33.9 30.6 43.9
qmax 26.8 29.5 34.3 36.3 35.8 34.1 32.0 32.0 32.3 31.8 29.5 27.0 31.8
qmin 13.6 16.5 21.5 25.0 26.5 26.7 26.3 26.3 26.1 23.9 18.4 14.2 22.1
13. Kolkata qmean 20.2 23.0 27.9 30.6 31.2 30.4 29.2 29.2 29.2 27.8 24.0 20.6 27.0
qa 25.2 28.0 32.9 35.6 36.2 35.4 34.2 34.2 34.2 32.8 29.0 25.6 32.0
% Rated
Load 106 103 99 96 96 97 98 98 98 99 102 105 100
qext 35.0 36.7 39.3 40.0 39.4 38.3 35.6 35.6 35.8 36.1 34.4 35.0 40.0
qmax 29.7 32.2 34.7 34.6 33.5 30.5 29.0 29.9 30.7 30.4 29.3 28.9 31.1
qmin 19.2 20.2 22.1 23.4 23.6 22.5 22.0 22.1 22.0 22.0 21.1 19.6 21.7
14. Coimbatore qmean 24.4 26.2 28.4 29.0 28.6 26.5 25.5 26.0 26.4 26.2 25.2 24.2 26.4
qa 29.4 31.2 33.4 34.0 33.6 31.5 30.5 31.0 31.4 31.2 30.2 29.2 31.4
% Rated
Load 102 101 99 98 98 100 101 101 101 101 102 102 101
qext 35.6 38.9 42.8 45.0 47.7 47.2 40.0 37.2 36.7 36.7 35.0 33.3 47.7
qmax 28.9 31.5 35.9 38.3 38.8 35.8 31.6 31.6 32.2 32.0 30.1 28.4 32.9
15. Cuttack qmin 15.7 18.2 22.1 25.3 26.9 26.5 25.6 25.6 25.5 23.7 18.8 15.5 22.5
qmean 22.3 24.8 29.0 31.8 32.8 31.2 31.2 28.6 28.8 27.8 24.4 22.0 27.7
qa 27.3 29.8 34.0 36.8 37.8 36.2 33.6 33.6 33.8 32.8 29.4 27.0 32.7
% Rated
Load 104 102 98 95 94 96 98 98 98 99 102 104 99
427
Sheet No. 8
428
qext 33.3 34.0 34.4 34.1 34.6 32.7 31.7 32.2 31.1 32.2 32.8 32.8 34.6
qmax 30.6 30.7 31.3 31.4 30.9 29.0 28.1 28.1 28.3 29.2 29.8 30.3 29.8
qmin 23.2 24.3 25.8 26.0 25.7 24.1 23.7 24.0 24.2 24.2 24.1 23.5 24.4
16. Cochin qmean 26.9 27.5 28.6 28.7 28.3 26.6 25.9 26.0 26.2 26.7 27.0 26.9 27.1
qa 31.9 32.5 33.6 33.7 33.6 31.6 30.9 31.0 31.2 31.7 32.0 31.9 32.1
% Rated
Load 100 100 98 98 99 100 101 101 101 100 100 100 100
qext 26.1 29.2 37.2 40.6 42.8 43.9 40.6 37.2 34.4 36.1 30.1 27.2 43.9
qmax 19.1 21.4 26.4 32.1 36.2 35.3 30.4 29.5 29.6 28.2 24.7 20.9 27.8
qmin 6.1 8.2 12.4 17.0 21.5 23.6 23.1 22.7 21.3 16.1 10.3 7.00 15.8
17. Dehradun qmean 12.6 14.8 19.4 24.6 28.8 29.4 26.8 26.1 25.4 22.2 17.5 14.0 21.8
qa 17.6 19.8 24.4 29.6 33.8 34.4 31.8 31.1 30.4 27.2 22.5 19.0 26.8
% Rated
Load 113 111 107 102 98 98 100 101 102 104 109 112 104
Transformers, 2/e
qext 32.2 38.3 40.6 43.9 46.1 46.1 36.7 36.1 35.6 35.6 32.8 30.6 46.1
qmax 24.7 27.4 33.2 37.6 38.8 35.9 31.2 30.8 31.2 30.6 27.8 25.2 31.1
qmin 12.1 14.5 19.2 23.5 25.9 25.9 24.8 24.8 24.3 21.7 15.6 12.4 20.4
18. Dhanbad qmean 18.4 21.0 26.2 30.6 32.4 30.9 28.0 27.8 27.8 26.2 21.7 18.8 25.8
qa 23.4 26.0 31.2 35.6 37.4 35.9 33.0 32.8 32.8 31.2 26.7 23.8 30.8
% Rated
Load 108 105 101 96 95 96 99 99 99 101 104 107 101
Sheet No. 9
qext 28.9 35.0 38.3 40.6 41.1 36.7 37.2 37.8 37.2 35.6 32.2 29.4 41.1
q max 24.0 26.3 30.2 31.6 31.0 31.5 32.1 32.2 32.1 30.5 27.7 24.9 29.5
qmin 11.0 12.8 16.5 20.3 22.7 24.7 25.8 25.8 25.2 22.0 16.9 12.5 19.7
19. Guahati qmean 17.5 19.6 23.4 26.0 26.9 28.1 29.0 29.0 28.6 26.2 22.3 18.7 24.6
qa 22.5 24.6 28.4 31.0 31.9 33.1 34.0 34.0 33.6 31.2 27.3 23.7 29.6
% Rated
Load 108 106 103 101 100 99 98 98 98 101 104 107 102
qext 30.6 34.4 45.6 47.9 48.3 47.8 47.2 43.3 42.2 41.7 36.7 33.6 48.3
q max 21.7 25.0 30.7 37.0 41.6 41.3 37.3 35.5 35.7 34.6 29.6 24.1 32.8
qmin 5.5 8.1 13.3 19.0 24.6 27.7 27.3 26.1 23.9 17.4 9.8 6.0 17.4
20. Hissar qmean 13.6 16.6 22.0 28.0 33.1 34.5 32.3 30.8 29.8 26.0 19.7 15.0 25.1
qa 18.6 21.6 27.0 33.0 38.1 39.5 37.3 35.8 34.8 31.0 24.7 20.0 30.1
% Rated
Load 112 109 104 99 94 92 95 96 97 101 106 111 102
qext 35.0 37.2 42.2 43.3 44.4 43.9 37.2 36.1 36.1 36.7 33.9 33.3 44.4
q max 28.6 31.2 34.8 36.9 38.7 34.1 29.8 29.5 29.7 30.3 28.7 27.8 31.7
qmin 14.6 16.7 20.0 23.7 26.2 24.1 22.3 22.1 21.6 19.8 16.0 13.4 20.0
21. Hyderabad qmean 21.6 24.0 27.4 30.3 32.4 29.1 26.0 25.8 25.6 25.0 22.4 20.6 25.8
qa 26.6 29.0 32.4 35.3 37.4 34.1 31.0 30.8 30.6 30.0 27.4 25.6 30.8
% Rated
Load 104 102 100 97 95 98 101 101 101 102 104 105 101
429
Sheet No. 10
430
qext 32.8 37.2 41.1 45.6 46.7 46.1 41.7 35.0 35.6 36.7 33.9 32.8 46.7
q max 26.1 28.9 34.0 38.5 41.9 37.6 30.3 29.5 30.8 31.4 28.9 26.9 32.1
qmin 9.8 11.4 15.5 20.5 25.9 26.4 23.9 23.6 23.1 18.4 11.7 9.0 18.3
22. Jabalpur qmean 18.0 20.2 24.8 29.5 33.9 32.0 27.1 26.6 27.0 24.9 20.3 18.0 25.2
qa 23.0 25.2 29.8 34.5 38.9 37.0 32.1 31.6 32.0 29.9 25.3 23.0 30.2
% Rated
Load 108 106 102 98 93 95 100 100 100 102 106 108 102
qext 31.7 36.7 42.8 44.9 47.8 47.2 46.7 41.7 41.7 40.0 36.1 31.1 47.8
q max 22.0 25.4 30.9 36.5 40.6 39.2 34.1 31.9 33.2 33.2 29.0 24.4 31.7
qmin 8.3 10.7 15.5 21.0 25.8 27.3 25.6 24.3 23.0 18.3 12.0 9.1 18.4
23. Jaipur qmean 15.2 18.0 23.2 28.8 33.2 33.2 29.8 28.1 28.1 25.8 20.5 16.8 25.0
qa 20.2 23.0 28.2 33.8 38.2 38.2 34.8 33.1 33.1 30.8 25.5 21.8 30.0
% Rated
Load 111 108 103 98 94 94 97 99 99 101 105 109 102
Transformers, 2/e
qext 26.7 31.7 37.2 43.9 46.1 47.2 45.0 41.7 38.3 37.2 32.5 27.2 47.2
q max 18.4 21.1 26.4 33.1 39.0 40.4 35.4 33.2 33.3 31.4 26.2 21.1 30.0
qmin 8.3 10.6 14.8 20.5 25.8 27.7 26.0 25.1 23.9 19.4 13.4 9.3 18.7
24. Jammu qmean 13.4 15.8 20.6 26.8 32.4 34.0 30.7 29.2 28.6 25.4 19.8 15.2 24.4
qa 18.4 20.8 25.6 31.8 37.4 39.0 35.7 34.2 33.6 30.4 24.8 20.2 29.4
% Rated
Load 113 110 105 100 95 93 96 98 98 102 106 111 102
Sheet No. 11
qext 33.3 37.8 43.3 45.6 47.2 47.8 45.6 42.2 40.6 40.6 36.1 32.8 47.8
q max 24.1 27.5 33.5 38.9 42.6 40.4 33.5 31.7 32.5 33.3 29.7 25.5 32.8
25. Jhansi qmin 9.2 11.7 17.4 23.3 28.8 29.3 25.9 24.9 24.1 19.5 13.1 9.3 19.7
qmean 16.6 19.6 25.4 31.1 35.7 34.8 29.7 27.9 28.3 26.4 21.4 17.4 26.2
qa 21.6 24.6 30.4 36.1 40.7 39.8 34.7 32.9 33.3 31.4 26.4 22.4 31.2
% Rated
Load 109 106 102 96 91 92 97 99 99 101 105 109 101
qext 25.0 28.3 33.3 37.2 37.5 37.8 32.8 33.3 33.3 33.3 29.4 28.3 37.8
q max 17.8 20.7 25.2 28.9 30.4 29.5 28.5 28.5 28.1 26.3 22.7 19.1 25.5
26. Kathmandu qmin 1.9 4.0 8.0 11.5 15.9 19.1 20.1 19.9 18.6 13.5 7.4 3.0 11.9
(Nepal) qmean 9.8 12.4 16.6 20.2 23.2 24.3 24.3 24.2 23.4 19.9 15.0 11.0 18.7
qa 14.8 17.4 21.6 25.2 28.2 29.3 29.3 29.2 28.4 24.9 20.0 16.0 23.7
% Rated
Load 116 114 109 106 103 102 102 102 103 106 111 115 107
qext 35.0 38.9 43.3 45.0 47.2 46.7 39.4 37.8 37.2 37.4 33.9 33.4 47.2
q max 31.0 33.6 36.8 39.0 41.3 37.6 32.6 32.2 32.6 32.5 30.6 30.1 34.2
qmin 17.6 20.0 23.1 25.9 28.1 27.2 24.9 24.7 24.4 22.9 19.1 16.7 22.9
27. Khammam qmean 24.3 26.8 30.0 32.4 34.7 32.4 28.8 28.4 28.5 27.7 24.8 23.4 28.6
qa 29.3 31.8 35.0 37.4 39.7 37.4 33.8 33.4 33.5 32.7 29.8 28.4 33.6
% Rated
Load 102 100 97 95 92 95 98 99 98 99 102 103 98
431
Sheet No. 12
432
qext 33.9 38.3 42.8 45.6 47.8 47.8 47.2 41.1 40.6 41.1 37.2 33.9 47.8
q max 24.5 28.5 34.1 39.0 42.6 40.3 33.3 31.7 33.1 34.5 30.8 26.7 33.3
28. Kota qmin 10.6 13.1 18.5 24.4 29.7 29.5 26.4 25.4 24.7 21.0 14.8 11.3 20.8
qmean 17.6 20.8 27.3 31.7 36.2 34.9 29.8 28.6 28.9 27.8 22.8 19.0 27.0
qa 22.6 25.8 32.3 36.7 41.2 39.9 34.8 33.6 33.9 32.8 27.8 24.0 32.0
% Rated
Load 108 105 100 95 91 92 97 98 98 99 103 107 100
qext 36.1 38.9 41.7 44.4 45.6 44.4 38.3 37.8 37.8 38.3 36.1 34.4 45.6
q max 31.3 34.3 37.5 39.3 40.0 35.6 32.5 32.1 31.9 32.4 31.0 30.3 34.0
29. Karnool qmin 17.0 19.3 22.5 26.0 27.2 25.0 23.8 23.5 23.3 22.4 19.2 16.6 22.1
qmean 24.2 26.8 30.0 32.6 33.6 30.3 28.2 28.5 28.3 27.4 24.2 21.6 27.1
qa 29.2 31.8 35.0 37.6 38.6 35.3 33.2 33.5 33.3 32.4 29.2 26.6 32.1
% Rated
Load 102 100 97 94 93 97 99 99 99 100 102 104 100
Transformers, 2/e
qext 30.6 35.0 41.7 45.6 47.2 48.3 45.6 38.9 39.4 40.0 35.0 33.3 48.3
q max 23.3 26.4 32.9 38.3 41.2 39.3 33.6 32.5 33.0 32.8 29.3 24.8 32.3
30. Lucknow qmin 8.9 11.5 16.3 21.8 26.5 28.0 26.6 26.0 25.1 19.8 12.7 9.1 19.4
qmean 16.1 19.0 24.6 30.1 33.8 33.6 30.1 29.2 29.0 26.3 21.0 17.0 25.8
qa 21.1 24.0 29.6 35.1 38.8 38.6 35.1 34.2 34.0 31.3 26.0 22.0 30.8
% Rated
Load 110 107 102 97 93 93 97 98 98 101 105 109 101
Sheet No. 13
qext 28.9 33.3 41.1 46.1 48.3 47.9 47.8 44.4 41.7 40.0 35.0 29.4 48.3
qmax 20.2 23.3 29.0 36.0 41.2 41.1 36.0 34.7 35.3 33.9 28.8 22.9 31.9
qmin 5.8 8.4 12.9 18.5 24.2 27.1 26.7 26.1 23.9 17.5 10.1 6.2 17.3
31. Ludhiana qmean 13.0 15.8 21.0 27.2 32.7 34.1 31.4 30.4 29.6 25.7 19.4 14.6 24.6
qa 18.0 20.8 26.0 32.2 37.7 39.1 36.4 35.4 34.6 30.7 24.4 19.6 29.6
% Rated
Load 113 110 105 100 94 93 96 97 97 101 107 111 102
qext 32.8 36.7 40.6 42.8 45.0 43.3 41.1 40.0 38.9 39.4 34.4 32.8 45.0
qmax 28.8 30.6 32.7 34.9 37.6 37.3 35.2 34.5 33.9 31.8 29.2 28.2 32.9
qmin 20.3 21.1 23.1 26.0 27.8 27.6 26.3 25.8 25.4 24.4 22.5 21.0 24.3
32. Madras qmean 24.6 25.8 27.9 30.4 32.7 32.3 30.8 30.2 29.6 28.1 25.8 24.6 28.6
qa 29.6 30.8 32.9 35.4 37.7 37.3 35.8 35.2 34.6 33.1 30.8 29.6 33.6
% Rated
Load 102 101 99 97 94 95 96 97 97 99 101 102 98
qext 34.4 38.3 41.7 41.7 41.7 42.2 40.6 40.0 39.4 38.3 36.1 35.0 42.2
qmax 30.2 32.4 35.0 36.3 37.5 36.7 35.7 35.3 35.0 33.0 30.6 29.7 33.9
33. Madurai qmin 20.9 21.6 23.4 25.4 26.3 26.3 25.7 25.2 24.8 24.0 23.0 21.6 24.0
qmean 25.6 27.0 29.2 30.8 31.9 31.5 30.7 30.2 29.9 28.5 26.8 25.6 29.0
qa 30.6 32.0 34.2 35.8 36.9 36.5 35.7 35.2 34.9 33.5 31.8 30.6 34.0
% Rated
Load 101 100 98 96 95 96 96 97 97 98 100 101 98
433
Sheet No. 14
434
qext 34.4 37.2 35.0 33.9 33.3 33.3 31.7 30.6 31.7 36.1 35.6 33.9 37.2
qmax 29.7 29.0 30.0 30.9 31.3 29.4 28.0 27.8 28.1 29.8 31.0 30.5 29.5
34. Marmugao qmin 21.4 21.9 23.9 26.1 29.9 24.7 24.0 23.9 23.8 23.9 22.8 21.5 23.7
qmean 25.6 25.4 27.0 28.5 29.1 27.0 26.0 25.8 26.0 26.8 26.9 26.0 26.6
qa 30.6 30.4 32.0 33.5 34.1 32.0 31.0 30.8 31.0 31.8 31.9 31.0 31.6
% Rated
Load 101 102 100 98 98 100 101 101 101 100 100 101 100
qext 32.8 36.1 37.8 39.4 37.8 37.2 33.3 33.9 33.3 32.8 32.2 31.7 39.4
qmax 28.3 31.2 33.5 34.0 32.6 28.9 27.3 27.9 28.7 28.4 27.4 27.0 29.6
35. Mysore qmin 16.4 18.2 20.2 21.4 21.2 20.2 19.7 19.6 19.3 19.6 18.3 16.5 19.2
qmean 22.4 24.7 26.8 27.7 26.9 24.6 23.5 23.8 24.0 24.0 22.8 21.8 24.4
qa 27.4 29.7 31.8 32.7 31.9 29.6 28.5 28.8 29.0 29.0 27.8 26.8 29.4
% Rated
Load 104 102 100 99 100 102 103 102 102 102 103 104 102
Transformers, 2/e
qext 35.0 38.9 45.0 46.1 47.8 47.2 40.6 37.8 38.9 38.3 35.6 33.9 47.8
qmax 28.6 32.5 36.4 39.7 42.8 38.4 31.2 30.4 31.5 31.9 29.9 28.7 33.5
36. Nagpur qmin 12.7 15.1 19.1 23.9 28.4 26.9 24.0 23.7 23.1 20.0 14.1 12.1 20.3
qmean 20.6 23.8 27.8 31.8 35.6 32.6 27.6 27.0 27.3 26.0 22.0 20.4 26.9
qa 25.6 28.8 32.8 36.8 40.6 37.6 32.6 32.0 32.3 33.0 27.0 25.4 31.9
% Rated
Load 105 102 99 95 91 94 99 100 100 99 104 106 100
Sheet No. 15
qext 29.4 33.3 40.6 45.6 47.2 46.7 45.0 40.0 40.6 39.4 35.0 28.9 47.2
qmax 21.3 23.6 30.2 36.2 40.5 39.9 35.3 33.7 34.1 33.1 28.7 23.4 31.7
37. New Delhi qmin 7.3 10.1 15.1 21.0 26.6 28.7 27.2 26.1 24.6 18.7 11.8 8.0 18.8
qmean 14.3 16.8 22.6 28.6 33.5 34.3 31.2 29.9 29.4 25.9 20.2 15.7 25.2
qa 19.3 21.8 27.6 33.6 38.5 39.3 36.2 34.9 34.4 30.9 25.2 20.7 30.2
% Rated
Load 112 109 103 98 94 93 96 97 98 101 106 110 102
qext 28.9 34.4 40.6 43.3 45.6 46.1 41.7 38.3 37.8 36.1 33.9 30.6 46.1
qmax 23.6 26.3 32.9 37.6 38.9 36.7 32.9 32.1 32.3 31.9 28.9 24.9 31.6
38. Patna qmin 11.0 13.4 18.6 23.3 26.0 27.1 26.7 26.6 26.3 23.0 16.1 11.7 20.8
qmean 17.3 19.8 25.8 30.4 32.4 31.9 29.8 29.4 29.3 27.4 22.5 18.3 26.2
qa 22.3 24.8 30.8 35.4 37.4 36.9 34.8 34.4 34.3 32.4 27.5 23.3 31.2
% Rated
Load 109 106 101 97 95 95 97 98 98 100 103 108 101
qext 35.0 38.9 42.8 43.3 43.3 41.7 35.6 33.0 36.1 37.8 36.1 35.0 43.3
qmax 30.7 32.9 36.1 37.9 37.2 31.9 27.8 27.7 29.2 31.8 30.8 30.1 32.0
39. Poona qmin 12.0 13.3 16.8 20.6 22.6 23.0 22.0 21.5 20.8 19.3 15.0 12.0 18.2
qmean 21.4 23.1 26.4 29.2 29.9 27.4 24.9 24.6 25.0 25.6 22.9 21.0 25.1
qa 26.4 28.1 31.4 34.2 34.9 32.4 29.9 29.6 30.0 30.6 27.9 26.0 30.1
% Rated
Load 105 103 101 98 97 100 102 102 102 101 103 105 102
435
Sheet No. 16
436
qext 32.8 33.0 40.0 41.1 42.2 39.4 36.7 36.7 36.1 36.1 33.9 32.8 42.2
40. Puri q max 26.9 28.3 30.0 30.7 31.6 31.7 30.6 31.0 31.4 31.2 29.3 27.2 30.0
qmin 17.9 20.8 24.6 26.6 27.7 27.4 26.7 26.8 26.6 25.0 20.8 17.7 24.1
qmean 22.4 24.6 27.3 28.6 29.6 29.6 28.6 28.9 29.0 28.1 25.0 22.4 27.0
qa 27.4 29.6 32.3 33.6 34.6 34.6 33.6 33.6 34.0 33.1 30.0 27.4 32.0
% Rated
Load 104 102 100 98 97 97 98 98 98 99 102 104 100
qext 33.9 38.9 42.8 44.6 47.2 47.2 40.0 37.4 37.3 36.4 35.2 34.1 47.2
q max 31.1 34.1 37.7 40.3 42.8 38.6 32.1 31.3 32.0 32.5 30.7 30.2 34.5
41. Ramgundam qmin 16.1 18.8 22.7 26.9 29.7 28.2 24.7 24.4 24.4 22.8 17.5 15.0 22.6
qmean 23.6 26.4 30.2 33.6 36.2 33.4 28.4 27.8 28.2 27.6 24.1 22.6 28.6
qa 28.6 31.4 35.2 38.6 41.2 38.4 33.4 32.8 33.2 32.6 29.1 27.6 33.6
% Rated
Load 102 101 97 93 91 94 99 99 99 99 102 103 98
Transformers, 2/e
qext 21.1 24.4 28.9 30.0 30.7 28.3 28.3 29.4 27.8 27.2 25.6 22.8 30.7
q max 15.5 17.1 21.5 23.8 23.7 23.7 24.1 24.1 23.6 21.8 18.9 16.4 21.2
42. Shillong qmin 3.6 6.4 10.5 14.1 15.5 17.4 18.1 17.8 16.6 12.9 7.7 4.5 12.1
qmean 9.6 11.8 16.0 19.0 19.6 20.6 21.1 21.0 20.1 17.4 13.3 10.4 16.6
qa 14.6 16.8 21.0 24.0 24.6 25.6 26.1 26.0 25.1 22.4 18.3 15.4 21.6
% Rated
Load 116 114 110 107 106 105 105 105 106 109 113 116 109
Sheet No. 17
qext 18.9 20.6 23.9 28.3 30.0 30.6 28.9 27.8 25.0 23.9 21.1 20.4 30.6
43. Shimla qmax 8.5 10.3 14.4 19.2 23.4 24.3 21.0 20.1 20.0 17.9 15.0 11.3 17.1
qmin 1.9 3.1 6.8 11.2 15.0 16.2 15.6 15.2 13.8 10.8 7.3 4.2 10.1
qmean 5.2 6.7 10.6 15.2 19.2 20.2 18.3 17.6 16.9 14.3 11.2 7.8 13.6
qa 10.2 11.7 15.6 20.2 24.2 25.2 23.3 22.6 21.9 19.3 16.2 12.8 18.6
% Rated
Load 120 118 115 111 107 106 108 108 109 112 115 117 112
qext 17.2 20.6 25.6 31.1 35.6 37.8 38.3 36.4 33.0 33.9 23.9 18.3 38.3
qmax 4.4 7.9 13.4 19.3 24.6 29.0 30.8 29.9 28.3 22.6 15.5 8.8 19.5
44. Srinagar qmin –2.3 –0.8 3.5 7.4 11.2 14.4 18.4 17.9 12.7 5.7 –0.1 –1.8 7.2
qmean 1.0 3.6 8.4 13.4 17.9 21.7 24.6 23.9 20.5 14.2 7.7 3.5 13.4
qa 6.0 8.6 13.4 18.4 22.9 26.7 29.6 28.9 25.5 19.2 12.7 8.5 18.4
% Rated
Load 124 122 177 133 108 104 102 102 105 112 117 122 113
qext 38.3 41.7 43.9 45.6 45.6 45.6 38.9 37.2 41.1 41.1 39.4 38.9 45.6
qmax 31.4 33.1 36.1 37.3 36.2 33.7 30.5 30.3 31.6 35.5 34.9 32.8 33.6
45. Surat 14.8 16.4 20.1 23.7 26.6 27.1 25.7 25.4 24.1 23.1 19.2 16.0 21.9
qmin
qmean 23.1 24.8 28.1 30.5 31.4 30.4 28.1 27.8 27.8 29.3 27.0 24.4 27.8
qa 28.1 29.8 33.1 35.5 36.4 35.4 33.1 32.8 32.8 34.3 32.0 29.4 32.8
% Rated
Load 103 102 99 96 96 97 99 99 99 98 100 102 99
437
Sheet No. 18
438
qext 35.6 40.0 42.2 42.8 43.3 43.9 41.1 40.6 40.6 36.9 36.7 35.6 43.9
q max 30.1 32.7 35.1 36.7 37.1 36.4 35.5 35.1 34.2 32.3 29.9 29.3 33.7
46. Tiruchy qmin 20.6 21.3 22.9 25.8 26.4 26.5 25.9 25.4 24.9 23.9 22.7 21.3 24.0
qmean 25.4 27.0 29.0 31.2 31.8 31.4 30.7 30.2 29.6 28.1 26.3 25.3 28.8
qa 30.4 32.0 34.0 36.2 36.8 36.4 35.7 35.2 34.6 33.1 31.3 30.3 33.8
% Rated
Load 102 100 98 96 95 95 95 97 97 99 101 102 98
qext 34.4 35.0 36.2 35.0 35.2 34.4 31.7 32.8 33.8 32.8 33.9 34.4 36.2
q max 31.3 31.7 32.5 32.4 31.6 29.4 29.1 29.4 29.9 29.9 30.1 30.9 30.7
qmin 22.3 22.9 24.2 25.1 25.0 23.6 23.2 23.3 23.3 23.4 23.1 22.5 23.5
47. Trivandrum qmean 26.8 27.3 28.4 28.8 28.3 26.5 26.2 26.4 26.6 26.6 26.6 26.7 27.2
qa 31.8 32.3 33.4 33.8 33.3 31.5 31.2 31.4 31.6 31.6 31.6 31.7 32.2
% Rated
Load 100 100 99 98 99 100 101 101 100 100 100 100 100
Transformers, 2/e
qext 33.1 36.7 38.3 40.5 43.3 44.4 38.3 38.2 37.8 36.8 33.9 32.8 44.4
q max 27.7 29.2 31.2 32.8 34.0 33.7 31.7 32.0 31.6 30.9 29.3 27.7 31.0
48. Vizag qmin 17.5 19.3 22.6 25.9 27.8 27.4 26.0 26.0 25.6 24.5 21.2 18.3 23.5
qmean 22.6 24.2 26.9 29.4 30.9 30.6 28.8 29.0 28.6 27.7 25.2 23.0 27.2
qa 27.6 29.2 31.9 34.4 35.9 35.6 33.8 34.0 33.6 32.7 30.2 28.0 32.2
% Rated
Load 103 102 100 98 96 96 98 98 98 99 102 103 100
C HAPTER 17
Erection and Commissioning

C.M. Shrivastava
The chapter describes some of the main precautions which must be

taken during erection and commissioning of a transformer. Mainte-
nance schedule has also been discussed so that continuous trouble-
free service could be ensured.
17.1 Despatch
After completing all contractual obligatory tests at works, trans-

former/reactor is made ready for despatch.
When transport conditions permit, transformers are despatched
— fully assembled including fittings
— assembled without fittings, with inert gas (preferably dry
nitrogen)
— assembled without fittings, with oil filled up to top yoke level
— partially assembled job for site assembly.
In a great majority of the cases accessories like radiators, bush-
ings, explosion vent/pressure relief valve, dehydrating breather,
rollers, Buchholz relay, conservator, pipe work, marshalling box are
despatched separately. In some of the cases, tapchanger is also des-
patched separately. Transformer oil (if it is included in the order) is
sent in separate sealed drums/tankers. When transformers are des-
patched with inert gas, positive pressure must be maintained
throughout the period till gas is replaced by oil.
17.2 Inspection upon Arrival at Site
Immediately after transformer is received at site, it should be thor-

oughly examined externally for possible damages which may have
occurred during transit. Nitrogen gas pressure (when filled during
despatch) should be checked. Positive pressure if not found indicates
that there is a leakage, and there is a possibility of the moisture
entering the tank during transit. This can be ascertained by measur-
ing the dew point. The dew point measurement indicates the amount
of surface moisture content in transformer insulation. As the insula-
tion temperature and transformer gas pressure vary, the acceptable
dew point will vary.
The various packages must also be checked. Internal inspection
should be carried out to the extent possible through inspection cov-
ers. Particular attention should be paid to the connections, bolts,
links, coil clamping bolts, tapchangers, current transformers and the
general insulation.
Breakdown strength of oil of transformer tank (when the trans-
former is despatched filled with oil) and drums containing trans-
former oil (which have been despatched separately) should be exam-
ined carefully.
17.3 Handling
Following means are normally used for lifting operations

— overhead travelling crane or gantry crane
— jib crane
— derricks
— jacks and winches
The overhead travelling crane and jib cranes are obviously the
most convenient and safe means. Precautions mentioned below
must be adhered to:
— Transformer (main package) should be lifted only through
lifting points provided for attaching the slings
— Cover must always be bolted in position
Erection and Commissioning 441
Transformer should be jacked up using the jacking pads specially

provided for that purpose. Jacks should never be placed under any
valves.
17.4 Installation
Following precautions should be taken before taking up erection

— Person going inside a transformer must wear clean clothes
and clean synthetic-rubber-soled sandals or boots.
— Never stand directly on any part of the insulation.
— No one should be allowed on top of the transformer, unless he
has emptied his pockets.
— All the tools and spanners used for erection should be
securely tied with taps so that these could be recovered if
dropped in, by accident.
— All components should be carefully cleaned outside separately,
before erecting.
Fibrous material should not be used for cleaning. The presence of
suspended fibrous material will reduce the electrical properties of
transformer oil.
— Interior of the transformer should not be exposed to damp
atmosphere as far as possible, to avoid condensation. In the
event of a sudden change in the weather bringing rain or
snow, provision must be made for closing the tank quickly and
pressurizing it with nitrogen so as to preserve the insulation.
— Naked lights and flame should never be used near oil filled
transformer. Smoking must not be allowed on the trans-
former cover when the cover plates are open, nor in the
vicinity of oil processing plant.
— Never allow any one to enter the transformer if adequate
supply of air in the tank is not available.
17.4.1 Location and Site Preparation
No special foundation is necessary for the installation of a trans-

former except a level floor strong enough to support the weight and
prevent accumulation of water. Foundation incorporating special oil

drainage facilities during fire and emergency is recommended for
large transformers. Transformers should be placed on the founda-
tion so that easy access is available all around and diagram plates,
thermometers, valves, oil gauges etc. can be easily reached or read.
Adequate electrical clearances are also to be provided from various
line points of the transformer to earthed parts.
Type “ONAN” transformer depends entirely upon the surrounding
air for carrying away the heat generated due to losses. For indoor
installation, therefore, the room must be well ventilated so that the
heated air can escape easily and be replaced by cool air. Air inlets
and outlets should be of sufficient size and numbers to pass adequate
air to cool the transformer. The inlets should be as near the floor as
possible and outlets as high as the building allows. Where necessary,
exhaust fans can be installed for the purpose. The transformer
should always be separated from one another and from all walls and
partitions to permit free circulation of air.
Where rollers are not fitted, level concrete plinth with bearing
plates of sufficient size and strength can be adopted for outdoor
transformers. The formation of rust, due to the presence of air and
water in the space between the plinth and the base of the trans-
former should be prevented by use of rust preventive bituminous
compound. Where rollers are fitted, suitable rails or tracks should
be used and the wheels should be locked to prevent accidental
movement of the transformer.
Bushing should be lifted properly as shown in Fig. 17.1. Mounting
of stress shield must be properly checked as per drawing and placed
accordingly. Lead should be pulled slowly so that joints should not
give way.
17.4.2 Special Precautions for Installing 245 kV Re-entrant

Type Bushing
The special feature of this bushing is the re-entrant control at the oil
end which is brought about by special arrangement of condenser
layers. Starting from the earth foil, the limiting area of foil ends in
conformity with regular constructional principle form a core oriented
away from the flange. At about two-third of the voltage between
flange and tube, the foil ends are reversed and follow a course of
Figure 17.1 Installation of bushing.

core reverting towards the flange and goes as far as the conductor
tube. The re-entrant control has the following advantages:
— The lower end is shorter than the conventional ends.
— Orifice of the lower end has a larger diameter and makes
possible introduction of thick insulated cable.
— Stress shield is avoided.
It is important to ensure the concentricity and axial position of the
insulated lead of transformer winding inside the re-entrant oil end
of the bushing. Following precautions must be followed (Refer to
Fig. 17.2):
— Ensure that the position of the transformer winding lead is
correct axially and radially.
Bushing
re-entrant
end
Incorrect lead Correct lead

position position
Metal hold
Figure 17.2 Handling of 245 kV re-entrant type bushing.

— Guide the lead while lowering the bushing on transformer

tank so that it is concentric with the bushing orifice and
angular gap all around is uniform.
— Check that lead is well insulated and in its final position the
insulation is extended to the inside of the bushing tube.
— Do not allow the bushing weight to fall on the lead insulation
while lowering the bushing into tank.
17.4.3 Oil Filling
Before filling with oil, transformer should be fitted with all accesso-
ries, such as valves, plugs, and made oil tight.
The oil which is to be filled in transformer must be tested for
dielectric strength and water content and should be in line with the
recommendations. If not, it must be filtered with stream line filters
with built-in heaters and vacuum pumps for improving the quality
of oil. During oil filling operation, it should be ensured that no air
pockets are left in the tank and no dust or moisture enters the oil
and it should be warmer than surrounding air.
For transformers despatched gas filled, the filling of oil inside the
tank should always be done under vacuum. While evacuating the
transformer tank, care should be taken to ensure that bakelite cylin-
ders, panels, etc., are not subjected to pressure. The vacuum should
be maintained for a short time after the tank is filled with oil.
The vacuum pulling in the tank will avoid chances of air getting
trapped and forming pockets. Oil filling will protect the windings
against exposure to atmosphere at the time of erection of bushings,
etc.
Following precautions should be taken during this operation:
— Oil is easily contaminated. It is very important, when
sampling the oil and filling the tank, to keep the oil free from
contamination.
— All equipment used for handling the oil should be cleaned and
flushed with clean transformer oil before use (the oil used for
washing must be discarded). Particular attention should be
paid to the cleanliness of bungs, valves and other parts where
dirt or moisture tends to collect.
— For sampling, glass containers with glass stoppers are to be
preferred over the metal type which are susceptible to
contamination by dirt, etc. Cleanliness is essential as even

small amount of dirt and water will affect the accuracy of test
results. Wax should not be used for sealing the oil sample
bottles. However, the stopper can be covered by a pack of
silicagel tied in a piece of cloth.
— Flexible steel hose is recommended for handling insulating
oil. Some kind of synthetic rubber or PVC hoses are also
suitable, but only those known to be satisfactory, should be
used. Ordinary rubber hose should not be used for this
purpose, as oil dissolves the sulphur from the rubber and is
thereby contaminated. Hose used for handing oil should be
clean and free from loose rust or scales.
— Transformer must always be disconnected from the
electricity supply system before the oil level in the tank is
lowered.
— Oil must not be emptied near naked lights as the vapour
released is inflammable.
17.4.4 Drying of Transformers
Drying of transformer is necessary in case insulation has absorbed

moisture.
The process of drying out a transformer is one requiring care and
good judgement. If the drying out process is improperly performed,
great damage may result to the transformer insulation through
overheating, etc. A properly dried out and correctly installed trans-
former is one of the most reliable electrical appliances. In no case
should a transformer be left unattended during any part of the dry-
out period; transformer should be carefully watched throughout the
dry-out process and all observations should be recorded properly.
Drying of Core and Coils Using Oven

Where a suitable oven is available, the core and coils can be effec-
tively dried in it by raising the temperature to a level not exceeding
80°C. A large volume of air should pass through the oven to remove
moisture and vapours. Insulation resistance check will indicate
when the coils are dry.
Core and coils can also be dried in its own tank in an oven. Trans-
former tank should be suitable for full vacuum (Low vacuum for low
voltage class). Full vacuum is kept in the tank and a temperature of
about 75°C is maintained. Dry nitrogen is used for breathing the

vacuum.
Drying by Short-circuit Method

The transformer can also be dried by heating the coils by short-
circuiting the low voltage winding and supplying a reduced voltage at
high voltage terminals. Current should not exceed 70% of normal
rated current and oil temperature should be of the order of 75°C.
Winding temperature should in no case exceed 90°C. The winding
temperature can be monitored by measuring winding resistance.
This method is more effective in drying the insulation at site.
Drying out by Streamline Filter Machine (Ref. Fig. 17.3)

The most practical method of drying out is by circulation of hot oil
through streamline filter machine incorporating oil heater and
vacuum chamber. The vacuum pump of the filter machine should
have the capacity of creating vacuum as high as possible but not less
than 710 mm of mercury. Drying out process can be made faster by
creating vacuum in the transformer tank by lagging the transformer
tank to prevent loss of heat. The oil temperature in transformer
should be of the order of 75°C.
It should be seen that the oil temperature at the filter machine in
no case exceeds 85°C.
Drying process can be terminated when transformer oil charac-
teristics are achieved within permissible limits and insulation resis-
tance of winding shows a constant or rising trend.
17.5 Commissioning
Transformer must be healthy in all respects before energizing and,

therefore, it should be thoroughly checked before commissioning.
Following checks should be carried out.
17.5.1 Transformer
Measurements
Measurement of ratio, resistance, vector group and magnetizing cur-
rent should be of the order of works’ test results. BDV and water
content of oil in transformer tank, tapchanger and cooler bank (the
Air gases and moisture being removed
from transformer
448
Moisture and dissolved gases

being evolved and taken off by
vacuum system
Ionic reaction Medium
coloumn (optional) Multistage vacuum
degassing system for
chamber degassing
chamber
Facility for
paralleling
vacuum
Edge filter
systems
Oil
tank
Moisture and
dissolved
Transformers, 2/e
gases being
evolved and
Path of oil Heater taken off by
flow tank vacuum system
Roots rotary
Transformer Discharge pump combination for
Gear pump taking oil Roots rotary
undergoing evacuating
from tank or transformer combination for
treatment transformer
high vacuum in
degassing
chamber
Figure 17.3 General diagram for treatment of oil and conditioning of transformer.
oil results should be within the permissible limits as mentioned in

Appendix A1 of IEC 422) have also to be measured.
Observations
Bushing CT polarity, silicagel breather, earthing, bushing arching
horn, valves, etc. must be checked for their correctness.
17.5.2 Protection
Buchholz relay, oil temperature, winding temperature, pressure

relief device, magnetic oil level gauge, differential over current,
earth fault, etc. Protection scheme should be checked for proper
functioning.
17.5.3 Tapchanger
Manually, electrical (local and remote), parallel operations, IR value

of motor and control wiring to be checked.
17.5.4 Marshalling Box
All control wiring should be checked.

When all the above checks have been done, a settling time of ap-
proximately 24 hr should be given to oil before charging the trans-
former. During this interval of settling, air should be released six
hourly wherever possible. Voltage should be slowly built up to full
level in around four to six hours wherever possible.
17.6 Maintenance
If a transformer is to give long and trouble-free service, it should

receive a reasonable amount of attention and maintenance. The
principal object of maintenance is to maintain the insulation in good
condition. Moisture, dirt and excessive heat are the main cause of
insulation deterioration. Maintenance consists of regular inspection,
testing and reconditioning, wherever necessary.
Load voltage, load current, temperature of winding, oil and
ambient should be recorded daily. Various fittings/accessories like
silicagel breather, bushing, tapchanger, fan pump, oil in tank,

tapchanger bushing, Buchholz relay, etc., should be checked periodi-
cally as per recommendation.
17.6.1 Reconditioning Procedure of Condenser Bushing

(S.R.B.P. Type)
In case moisture is found in the bushing, the oil of the bushing is

drained through drain plug. Two nylon tubes of suitable diameter
are connected to an oil filtering plant, through suitable adaptation.
The outlet tube from filter to be connected to the drain plug and the
inlet tube of the filter to be submerged in oil at the top. The filtered
hot oil should be circulated through bushing for three days by inject-
ing it at the bottom and sucking from the top. This not only improves
the oil quality but also dries out the bushing insulation. During the
entire period of circulation, the bushing top must be covered so that
no foreign matter can enter the bushing. After the reconditioning,
the IR value of the bushing and dielectric strength of oil should be in
line with the recommendations.
17.6.2 Analysis of Gases as a Means of Monitoring

Transformers and Detecting Incipient Fault
Gases are produced in a transformer when transformer oil is

subjected to high electrical or thermal stresses or due to breakdown
of other insulation materials. These gases get dissolved in oil and if
produced in substantial quantities, it may get collected in Buchholz
relay. Analysis of these gases, together with the rate of their forma-
tion makes it possible to estimate the nature and seriousness of the
fault.
The gases in Buchholz relay will, in general, be due to one of fol-
lowing three main causes:
(a) Air introduced because of mechanical fault in the oil system
or in some ancillary equipment.
(b) Gas produced by thermal or electrical breakdown of oil
without damage to any solid insulants.
(c) Gas produced by a thermal or electrical breakdown of solid
insulants which invariably includes the breakdown of oil.
It should, however, be kept in mind that one of the primary ob-

jects of fitting a Buchholz relay is to detect faults in their incipient
stage and care should be exercised in interpreting the analysis re-
sults together with all other available data, to ensure that re-ener-
gizing the transformer does not cause an incipient fault to develop
into a major fault.
In the event of low energy faults with only slight gas evolution,
and in view of the fact that the oil solubility of the gases of decompo-
sition may be very high, the detection of such gradual process
is delayed, because the Buchholz relay does not respond until a
sufficient volume of gas is collected. Further, because of the different
solubility of the various gas constituents and the partial exchange
with other gases dissolved in the oil, e.g. nitrogen or air, the compo-
sition of the gas mixture that is collected in the Buchholz relay
differs from that at the point of origin. It is, therefore, hardly
possible to obtain correct diagnosis of the nature of such an incipient
fault by simply analyzing the gas collected in the Buchholz relay. A
complete and reliable explanation of the faults can only be obtained
by supplementing the gas analysis of the Buchholz relay by an
analysis of the gases dissolved in the oil.
An analysis of the gas dissolved in the oil is possible by means of
gas thromatography, which determines the individual constituents
of such gas mixtures with a high degree of accuracy. IEC 599 Inter-
pretation of the analysis of gas in transformers and other oil filled
electrical equipment in service may be referred to for this analysis. A
comparison between the gases accumulated in the Buchholz relay
and those dissolved in the oil can be very useful in diagnosing the
nature and severity of the particular fault.
17.7 Dos for Power Transformer
1. Connect gas cylinder with automatic regulator if transformer

is to be stored for long duration, in order to maintain positive
pressure.
2. Fill the oil in the transformer at the earliest opportunity at
site and follow storage instructions. It must be commissioned
as soon as possible.
3. Open the equalizing valve between tank and OLTC diverter

compartment, whenever provided, at the time of filling the oil
in the tank and close the same during operation.
4. Clean the oil conservator thoroughly before erecting.
5. Check the pointers of all gauges for their free movement
before erection.
6. Inspect the painting and if necessary do retouching.
7. If inspection covers are opened or any gasket joint is
tightened, tighten the bolts evenly with the proper sequence
to avoid uneven pressure.
8. Clean the Buchholz relay and check the operation of alarm
and trip contacts.
9. Check the oil level in oil cup and ensure that the air passages
are free in the breather. If oil is less, make up the oil level.
10. Check the oil in transformer and OLTC for dielectric strength
and moisture content, and take suitable action for restoring
the quality of oil.
11. Attend to leakages on the bushing immediately.
12. Check the diaphragm of the relief vent. If cracked or broken,
replace it.
13. Remove the air from vent plug of the diverter switch before
energizing the transformer.
14. Check the gear box oil level in the tapchanger. If less, top up
with specified oil.
15. Check the OTI and WTI pockets and replenish the oil, if
required.
16. Check the oil level in the diverter switch and if found less, top
up with fresh oil.
17. Examine the diverter and selector contacts of tapchanger and
if found burnt or worn out, replace the same.
18. Check and thoroughly investigate the transformer whenever
any alarm or protection is operated.
19. Examine the bushings for dirt deposits and coats and clean
them periodically.
20. Check the protection circuits periodically.
21. Check all bearings and operating mechanism of the tapchanger
and lubricate them as per schedule.
22. Keep the valve connected between conservator of the

tap-changer and its diverter compartment open, during
transformer operation.
23. Check the silicagel charge. If it is found pink, regenerate or
replace it with blue silicagel charge.
17.8 Don’ts for Power Transformers
1. Do not use low capacity lifting jacks on transformer for

jacking.
2. Do not allow WTI, OTI temperature to exceed 75°C during
dryout of transformer, and filter machine temperature
beyond 85°C.
3. Do not re-energize the transformer, unless the Buchholz gas
is analyzed.
4. Do not re-energize the transformer without conducting all
pre-commissioning checks.
5. Do not energize the transformer, unless the off-circuit tap
switch handle is in locked position.
6. Do not leave off-circuit tap switch handle unlocked.
7. Do not leave tertiary terminals unprotected outside the tank.
Follow manufacturer’s recommendations in this regard.
8. Do not leave marshalling box doors open. They must be locked.
9. Do not leave any connection loose.
10. Do not meddle with the protection circuits.
11. Do not leave maximum temperature indicating pointer
behind the other pointer in OTI and WTI.
12. Do not change the settings of WTI and OTI alarm and
trip frequently. The setting should be done as per the site
conditions.
13. Do not allow oil level in the bushings to fall; they must imme-
diately be topped up.
14. Do not allow conservator oil level to fall below one-fourth level.
15. Do not parallel transformers which do not fulfil the required
conditions.
16. Do not switch off the heater in marshalling box except in
summer.
17. Do not leave secondary terminals on an unloaded CT open.

18. Do not allow water pressure more than oil pressure in
differential pressure gauge in OFWF cooled transformer.
19. Do not switch on water pump unless oil pump is switched on.
20. Do not leave ladder unlocked, when the transformer is
energized.
21. Do not allow unauthorized entry near the transformer.
22. Do not overload the transformer other than the specified
limits mentioned in national/international standards.
23. Do not allow inferior oil to continue in transformer.
24. Do not handle the off-circuit tap switch when the transformer
is energized.
17.9 Dos and Don’ts for HV Condenser Bushings
17.9.1 Dos
1. Check the packing externally for possible transit damage

before unpacking.
2. Do unpacking with care to avoid any direct blow on bushing
or porcelain insulator.
3. Store the bushing in a shed or covered with tarpaulin to
protect it from moisture and rains. If removed from the crate,
keep it indoors with lower end protectives intact.
4. Handle the bushing with manila rope slings without any
undue force on porcelain insulator.
5. Clean the porcelain insulator thoroughly.
6. Remove the wax tape protection on the oil end of the bushing
(at the time of erection) and clean the surface with hot trans-
former oil.
7. Check the oil level and IR value of the bushing in vertical
position only, taking care that the bushing is cleaned and no
rope or sling etc., is touching the terminal and ground.
8. Check the breakdown value (BDV) of oil taken from drain plug
or siphoning from the bottom-most portion of bushing. This
should not be less than the recommended value.
9. Check the IR value and tan d value (if possible) with bushing
in position on transformer with jumper connection removed.
Record these readings for reference and guidance for future
measurements.
10. Check BDV of oil and IR value of each bushing periodically
during maintenance shut down. These values should be
comparable with the values recorded at the time of commis-
sioning.
11. Ensure to allow the air to escape from central tube to the
atmosphere while filling the transformer tank.
12. Maintain the log book records of periodical checks (i.e. tan d
and BDV of oil) up to date.
17.9.2 Don’ts
1. Do not unpack the bushing from the crate unless required to

be mounted on the transformer.
2. Do not remove the waxed tape protection/metal protective
hood from the oil end portion unless bushing is required for
use.
3. Do not store the bushing outdoors without any protective
covering.
4. Do not measure the IR value and tan d value without
thoroughly cleaning the porcelain and oil end portion.
5. Do not store the bushing without oil in porcelain.
6. Do not keep the top cap cover open for any longer time than
required as it contaminates the oil.
7. Do not tighten the nuts and bolts in excess to stop any
leakage, this could damage the cemented joints on porcelain.
C HAPTER 18
Transformer Protection
B.L. Rawat
Transformer being a vital equipment, its protection is equally impor-

tant. The subject of transformer protection can be categorized under
two major headings.
(a) Protection of the transformer against the effects of faults
occurring on any part of the system beyond the transformer.
(b) Protection against the effects of faults arising in the trans-
former.
18.1 Protection against External Faults
18.1.1 Short-Circuit
Short-circuit across any two or all the three lines may occur in the
system. The over currents produced because of such faults depend
upon system MVA feeding the fault, the voltage which has been
short-circuited and upon the impedance of the circuit up to the fault.
System short-circuits produce a relatively intense rate of heating of
the feeding transformer, as the copper-loss increases in proportion
to the square of per-unit fault current. The duration of external
short-circuit that a transformer shall sustain without damage, if the
current is limited only by self-impedance, is 2 s as per IS : 2026.
Large fault currents produce severe mechanical stresses in the
transformer; the maximum stress occurs during the first cycle of
asymmetrical fault current and so cannot be averted by automatic
tripping of the circuit. The control of such stress is therefore a
Transformer Protection 457
matter of transformer design. The transformers connected to a large

power system have their windings very securely braced in order to
minimize the effects of the mechanical forces to which they may be
subjected, due to short-circuit. IS : 2026 specifies typical values of
short-circuit impedance, for different ratings of transformers, to
limit the short-circuit current.
18.1.2 High Voltage Disturbances
High voltage disturbances are of two kinds

(a) Transient surge voltages
(b) Power frequency voltages
(a) Transient Surge Voltage

High voltage high frequency surge may arise in the system due to
any of the following:
(i) Arcing grounds if neutral point is isolated
(ii) Switching operation
(iii) Atmospheric disturbances
These disturbances principally take the form of travelling waves
having high amplitudes and steep wave fronts. On account of both
their high amplitudes and frequencies these surges may, upon reach-
ing the winding of transformer, breakdown the insulation between
turns adjacent to line terminal, causing short-circuits between turns
and producing extensive damage to the transformer winding. The
effects of these surge voltages may, however, largely be minimized
by designing the winding to withstand the application of a specified
surge test voltage and then ensuring that this test voltage is not
exceeded in service by the provision of a suitable surge diverter
mounted adjacent to the transformer terminals.
All types of surge diverters aim at attaining the same results, viz.
of shunting disturbing surges from the lines to earth to prevent their
reaching the transformer. In essence, the different kinds of valve-
type surge diverters employ several spark gaps in series with a non-
linear resistor. These non-linear resistors offer a low resistance path
to high voltage surge waves and hence these disturbances are dis-
charged to earth through the diverter. As the surge voltage falls, the
diverter resistance automatically increases and prevents the flow of

power current to earth. Figure 18.1 shows schematically how the
various dielectric paths of a three-phase transformer should be
shunted by surge diverters in order to protect the transformer
bushings and windings against surge voltages. A surge protective
device should have the following qualities:
Tank
Core
LV HV
Surge
diverter
Figure 18.1 Schematic diagram showing connection of surge diverters.
(i) Rapid response to impulse over voltage

(ii) Independance of wave polarity
(iii) Non-linear characteristic
(iv) High thermal capacity
(v) High system follow current interrupting capacity
(vi) Consistent characteristic
Now, zinc oxide arrestors having high reliability and better char-
acteristic are also available in the market.
(b) Power Frequency Over-voltage

Power frequency over-voltage causes an increase in the stress on the
insulation and a proportionate increase in the working flux. The lat-
ter effect causes an increase in the iron-loss and a disproportionately
large increase in magnetizing current. In addition, flux is diverted
from the laminated core structure into steel structural parts. In par-
ticular, under conditions of over-excitation of the core, the core bolts
which normally carry little flux, may be subjected to a large compo-
nent of flux diverted from the saturated region of core alongside.
Under such conditions, the bolts may be rapidly heated to a tem-
perature which destroys their own insulation and will damage the
coil insulation if the condition continues.
Reduction of frequency has an effect, with regard to flux density,

similar to that of over-voltage.
The overfluxing protection does not call for high speed tripping.
Instantaneous operation is undesirable, as this would cause tripping
on momentary system disturbances which can be born safely, but
normal condition must be restored or the transformer must be iso-
lated before significant damage is done to insulation structure. The
fundamental equation for the generation of emf in a transformer can
be arranged to give
E
j =k
f
It is necessary to detect a ratio E/f exceeding unity, E and f being
expressed in per-unit values of rated quantities. An overfluxing relay
whose E/f characteristic closely matches with that of the transformer
should be used to give alarm and signal to correct the disturbance. If
the condition persists for a long time, the transformer should be
disconnected from the system to protect it from severe damage.
18.2 Protection against Internal Faults
Considering next the means to be adopted for protecting the trans-

former against the effect of faults arising in the transformer, the
principal faults which occur are:
(i) Breakdown to earth–either of winding or terminal gear
(ii) Phase-to-phase fault
(iii) Inter-turn fault
(iv) Core faults
18.2.1 Earth Faults

(a) Star-connected winding with neutral point earthed through an
impedance. The fault current for the fault shown in Fig. 18.2 is
dependent on the value of earthing impedance and is also proportional
to the distance of the fault from neutral point as voltage at the fault
point will be directly proportional to this distance. The fault current
in the primary winding will depend on transformation ratio between
primary winding the short-circuited turns, which varies with the
position of fault in the winding.
From Fig. 18.2, it may be noted that the fault in the lower third
of the winding produces very little current through the primary
terminals.
If If
single phase earth fault current
Percentage of respective max.
100
Fault
current
If
Primary current
100
0
Distance of fault from neutral
(percentage of WDG)
Figure 18.2 Earth fault current in resistance earthed star winding.
(b) Star-connected winding with neutral point solidly earthed. In

this case, fault current is controlled by the leakage reactance of the
winding which varies in a complex manner with the position of fault
on the winding. As in the ear-
20
lier case, the voltage available
for fault current varies with
Current (per unit)
15
the position of fault on the
Fault current
winding. It is seen that the 10
reactance decreases very
rapidly for the fault point 5
approaching the neutral and
Primary current
hence the fault current is
0 20 40 60 80 100
highest for a fault near the
Distance of fault from neutral
neutral end of the winding. (percentage of winding)
The variation of current with
fault position is shown in Figure 18.3 Earth fault current in case
Fig. 18.3. of solidly earthed neutral.
From Fig. 18.3 it can be noticed that the fault current magnitude
stays very high throughout the winding. Also, after transformation
ratio the input current curve remains at a substantial level for faults
at most points along the winding.
(c) Restricted earth fault protection. A simple over-current and earth
fault system will not give good protection cover for a star-connected
primary winding, particularly if the neutral is earthed through an
impedance. The restricted earth fault protection schematic shown in
Fig. 18.4 improve the degree of protection very much. This scheme
is operative for faults on the star winding of the transformer. The
system will remain stable for all faults outside the zone. As whole
fault current is measured, a good degree of protection of the winding
is achieved even in the case of neutral being earthed through imped-
ance. In the case of solidly earthed neutral, fault current remains at
a very high value even to the last turn of the winding and hence
complete cover for earth fault is obtained with restricted earth fault
protection scheme.
U
V
W
Impedance
R
High impedance relay
Figure 18.4 Restricted earth fault protection for a star-connected winding.
18.2.2 Phase-to-phase Fault

Phase-to-phase faults in the transformer are rare. If such a fault
does occur, it will give rise to substantial current to operate instanta-
neous over-current relay on the primary side as well as the differen-
tial relay.
18.2.3 Inter-turn Fault

A high voltage transformer connected to an overhead transmission
system is very likely to be subjected to steep fronted impulse volt-
ages. A line surge which may be of a magnitude several times the
rated system voltage will concentrate on the end turns of the wind-
ing, because of the high equivalent frequency of the surge front. Also
LV winding is stressed because of the transferred surge voltage. It is
reported that a very high percentage of all transformer failures arise
from faults between turns. Inter-turn fault may also occur because of
mechanical forces on the winding due to external short-circuit.
Though there may be high short-circuit current between few turns
loop, the terminal current will be very small because of the high
ratio of transformation between the whole winding and the short-
circuited turns. If this turn-to-turn fault is not detected in the
earliest stage, the subsequent progress may destroy the evidence of
the true cause.
(a) Diffferential Protection

A differential scheme can be arranged to cover the complete trans-
former protection. This is possible because of the high efficiency of
transformer operation and the close equivalence of ampere-turns
balance on both primary and secondary windings. The rated currents
of primary and secondary windings differ in inverse ratio to the cor-
responding voltages. Therefore, current transformers should have
their primary rating to match the rated currents of the transformer
windings to which they are applied. To correct phase shift of current
because of star/delta-connection of transformer winding, the current
transformers secondaries should be connected in delta and star as
shown in Fig. 18.5.
For correcting voltage variation, almost all transformers are pro-
vided with tapping, which in turn change the ratio from the mean.
This will create an imbalance proportional to the ratio change. At
Bias winding of relay
Operating WDG
of relay
Figure 18.5 Schematic diagram of differential protection scheme.

maximum through-fault current, the spill output produced by the

small percentage unbalance may be substantial. Therefore, differen-
tial protection should be provided with a proportional bias of an
amount which exceeds in effect the maximum ratio deviation. This
stabilizes the protection under through-fault condition while still
permitting the system to have good basic sensitivity.
(b) Magnetizing Inrush Current Phenomenon

The magnetizing inrush phenomenon produces current input to
the primary winding, which has no equivalent on the secondary
side. The whole of the magnetizing current appears, therefore, as
unbalances. The normal bias for the inrush current is not effective
and increase of protection setting to a value (which would avoid
operation of relay) would make the protection of little value.
As the magnetizing inrush current phenomenon is transient,
stability can be achieved by providing a small time delay. This can be
achieved by various means. A kick fuse can be connected as shunt to
an instantaneous relay. This fuse is so rated that it carries inrush
current without blowing. However, for internal fault the fuse blows
and permits the relay to operate. An induction pattern relay may
have both a suitable time characteristic and also a through bias
feature. Two induction electromagnets operate on a single disc to
produce opposing torque.
In case of severe inrush current, the set time delay might be
insufficient to give stability. Also, to minimize the damage to impor-
tant transformers, it may be essential to clear the fault without
delay. For these cases, another solution to the inrush phenomenon
must be found. It can be noticed that the waveform of inrush current
and zone fault current differs greatly. This distortion of waveform
can be used to distinguish between the conditions. The differential
current is passed through a filter which extracts the second harmonic
current. This component is then applied to produce a restraining
quantity sufficient to overcome the operating tendency due to whole
of the inrush current which flows in the operating circuit.
(c) Combined Differential and Restricted Earth Fault Scheme

A set of additional phase correcting auxiliary current transformers
are required as shown in Fig. 18.6.
18.2.4 Core Faults

If any portion of the core insulation becomes defective or the lami-
nated structure of the core is bridged by any conducting material
Connecting
auxiliary
current
transformer Restricted earth fault
relay bias winding
Differential relay operating coil
Figure 18.6 Schematic diagram of differential and restricted

earth fault scheme.
which can permit sufficient eddy current to flow, it will cause serious
over-heating. The insulated core bolts are used for tightening the
core. If the insulation of these bolts fails and provided easy path for
eddy current, this will lead to over-heating. This additional core-loss,
although it causes severe local heating, will not produce a noticeable
change in the input current and cannot be detected by the normal
electrical protection. However, it is desirable to detect over-heating
condition before a major fault has been created. Excessive heating of
core will breakdown transformer oil with evolution of gas. This gas
rises to the conservator.
(a) All faults in transformer core and windings result in the
localized heating and breakdown of oil. When the fault is of
very minor type such as hot joint, gas is released slowly and
rises towards conservator. A major fault where severe arcing
takes place, causes rapid release of large volume of gas and
oil vapour. This violent evolution of gas and oil vapour does
not have time to escape and instead builds up pressure and
bodily displaces the oil, causing surge of oil to pass up the
relief pipe to the conservator. Recognition of the above action
by Buchholz, and the limitation of other means of detecting
certain types of fault, led to the development of a protective
device known by his name. Construction of Buchholz relay
has been described in Chapter 11. Two electrical contacts are
available in the Buchholz relay, one for alarm and the other
for trip. When the generation of gas is slow, the gas while
moving towards the conservator gets trapped in the relay and
displaces the oil. After a certain amount of oil is displaced the

alarm contact is made and gives signal in the control room
necessitating further investigation.
A violent action described above causes a surge of oil which
makes the trip contact and the transformer is isolated from
the circuit.
Because of its universal response to faults within the
transformer, some of which are difficult to detect by other
means, the Buchholz relay is invaluable.
(b) Because of heavy arcing inside transformer, if excessive
pressure is generated, it is released through the pressure
relief device. Also one electrical contact on the device is made
which trips the circuit breakers and isolated the transformer
from electrical circuit. Construction of the device has been
described in Chapter 11.
(c) The overload or persistent fault increases oil temperature or
winding temperature or both. If the temperature increases
beyond safe working limit, the electrical contacts provided on
these devices are made. When first contact is made it gives
alarm signal in the control room necessitating investigation.
If temperature rises further, another contact is made which
trips the transformer. Construction of these devices have
been described in Chapter 11.
REFERENCES
1. Austen Stigant, S. and A.C. Franklin, The J and P Transformer Book,
Newness-Butterworth London.
2. Protective Relay Application and Guide, General Electric Co. (U.K.).
C HAPTER 19
Reactors
C.M. Shrivastava
S.K. Mahajan
Reactors find a number of applications in the transmission and distri-

bution network as well as industrial plants. They are usually classi-
fied according to duty application, viz. current limiting, neutral
earthing, shunt, damping, tuning or filter, arc-suppression, smooth-
ing, etc. These reactors have typical characteristic requirements and
call for different constructions, viz. with air core or with gapped iron
core, with or without magnetic shield, for indoor or outdoor installa-
tion, for fixed and variable reactance, with or without taps, etc.
19.1 Series Reactors
These reactors are intended for series connection in a system, either

for limiting the current under system fault condition or load sharing
in parallel circuits.
19.1.1 Types
Series reactors are classified by their location or application in a

system. Some of the more common types are described hereafter.
(i) Generator Line Reactor

This is usually used with a generator to reduce stresses under 3-
phase short-circuit.
(ii) Feeder Reactor

This is located on feeders from operating stations or on sub-station
buses and is used to minimise the resultant effects of a short-circuit
on other parts of the system. This application often allows a lighter
Reactors 467
Generator Generator
Reactor
Reactor Reactor Reactor
Figure 19.1 Generator reactor. Figure 19.2 Feeder reactor.
construction of feeder circuit which can result in appreciable

cost saving due to their association with circuit breakers of smaller
capacities.
(iii) Duplex Reactor

It is a centre tapped reactor used as an effective device in dividing
current source and limiting the magnitude of fault current.
It is usually applicable where:
(a) There is a heavy concentration of load as in a power house.
(b) Local generation is present.
(c) The utilisation and the generation voltages are same.
(d) There is a large amount of load distribution at the same
voltage.
Generator
Reactor
Figure 19.3 Duplex reactor.
(iv) Tie Line Reactor

It provides an easier method of expanding a system at the existing
voltage. It has following advantages:
(a) Existing switchgear can be used with a little modification.
(b) It is a relatively low cost means of expanding a system.
Generator
Reactor
Existing load New load
Figure 19.4 Tie line reactor.
(v) Synchronizing Reactor

This reactor provides a method for expanding a system when the
utilization voltage and generated voltage are the same and it is not
feasible to go to higher voltage. This is done by sectionalizing the
existing bus and tying all the bus sections through reactors to a com-
mon bus. By this method interrupting duty is greatly reduced on the
existing breaker and new loads and additional incoming capacity
may be added. This arrangement is frequently found in industrial
power plants on very large system.
G G G G
New load
added
Reactor Reactor Reactor Reactor
Figure 19.5 Synchronizing reactor.
(vi) Motor Starting Reactor

This is used when it is necessary to reduce the starting current of
AC motors. This decreases stresses on the motor and reduces the
system dusturbances.
Reactors 469
Reactor
M 3f
supply
Arrangement No. 1
Reactor
3f
M
supply
Arrangement No. 2
Figure 19.6 Starting reactor.
(vii) Reactor for Capacitor Bank

This reactor is used to limit high transient inrush current flowing
into the capacitor bank when capacitor bank is switched in. This also
helps in suppressing harmonics present in the system. This is also
called damping reactor. This is connected in series with the capaci-
tors and is rated for highest inrush current and continuous current.
Supply Supply
Reactor
Reactor
Figure 19.7 Reactor with capacitor bank.
(viii) Smoothing Reactor

This is employed to filter out all harmonics present in DC power
system. This can be both HV DC smoothing reactor and low voltage
A.C. switch yard Valve D.C. switch yard
Reactor
Converter
transformer
Filter
Figure 19.8 Smoothing reactor.
DC smoothing reactor with large superimposed harmonic compo-

nents.
(ix) Line Trap

This is used on power lines having carrier channels for equipment
control and communication purposes and as such is a resonated
inductance coil. It helps in blocking out unwanted frequencies and
noise from the carrier channels.
Line trap
EHV transmission line
400 km approx
Figure 19.9 Line trap.
19.1.2 Calculation Data
If it is desired to reduce fault from Z to Y MVA by connecting the

reactor in series
Xi = U2
F 1 - 1I
HY Z K
Reactors 471
A ¥ 1000
I=
3U
Y ¥ 1000
IS =
3U
Q = I ¥ X1 ¥ 10–3
2
where U = System voltage (kV)

A = System throughput power (MVA)
I = Reactor current (amperes)
Xi = Reactor reactance (ohms)
Q = Rating (kVAR per phase)
Is = Short time current (kilo-amperes)
19.1.3 General Consideration/Characteristics
(i) Linear Characteristic

The reactor should have linear characteristic which means a straight
line relationship between its current and voltage. Such a character-
istic helps in limiting current under system fault conditions. The
advantages of this characteristics are mentioned below:
(a) Expensive equipments are protected from failure due to
excessive mechanical stresses and heating caused by short-
circuits.
(b) Service on other parts of the system is not interrupted during
a fault, since the main bus voltage is not appreciably reduced.
(c) There is a reduction in the current interruption duty of circuit
breaker, thereby permitting the installation of circuit breaker
of reduced rating.
However, it may be noted that in case of starting reactor it is not
necessary to have a linear characteristic as it is required for shorter
duration and not for limiting the short-circuit current.
(ii) Ability to Withstand Short-Circuit

Since the reactor is basically required to limit short-circuit current,
mechanical forces (radial and compressive forces) are taken care
of during design by selecting proper parameters of reactor and
by proper clamping of the winding. Reactor must also be able to
withstand short-circuit current from thermal point of view. The
maximum permissible value of temperature of the winding during a
short-circuit is given below as per IS : 2026.
Class of Temperature°C
Type Insulation Copper Aluminium
Oil A 250 200
Dry A 180 180
Dry E 250 200
Dry B 350 200
Dry F and H 350 —
(iii) Fully Insulated Winding

Series reactor is connected in series in the system and therefore
both the ends should be suitable to withstand lightning impulse volt-
age. This can be achieved by having fully insulated winding.
(iv) Other Requirements

Rated continuous current and rated short-time currents are usually
specified for such reactors. The rated impedance, in ohms per phase,
for such a reactor, usually pertains to short-time current at rated
frequency. Rated impedance in such a case is determined from the
oscillogram records of the steady-state value of test voltage and cur-
rent during demonstration of dynamic ability in accordance with IS :
2026 (Part I)-1977. This is classified as a special test. The routine test
for impedance measurement is done at continuous current, if appli-
cable. In case of reactor without magnetic shield, this test also veri-
fies the rated impedance (at short-time high current). Measurement
of loss applies only on reactor where continuous current is specified.
19.1.4 Construction Features

Series reactors are manufactured in various way, some of them are
described below:
(i) Oil Immersed Coreless Reactor

Such a reactor can be manufactured for either indoor or outdoor
services and in general for oil immersed type, the reactor and its
external appearance, fittings and accessories closely follow estab-
lished power transformer practice. A high proportion of the magnetic
flux due to the current in a reactor winding passes through the coil
itself. This flux would produce local eddy losses particularly at the
coil ends and could possibly cause local hot spots. To obviate the
above condition it is preferable to use thin stranded conductor and
frequent transposition in the coils. Transposed cables are also
Reactors 473
sometimes preferred. Due to heavy concentration of flux in and

around the coil, non-magnetic materials are used except main tank
body. To achieve constant ohmic reactance during short-circuit
conditions, usually coreless construction with non-magnetic shield-
ing of winding is provided. The non-magnetic shield (generally made
of aluminium or copper) is in the form of cylinder and located around
the winding. Eddy currents are induced in the shield, which not only
help in reducing the tank losses but also reduced any tendency for
the coil to bulge out during short-circuit.
(ii) Dry Type Coreless Reactor

Dry type reactor is usually cooled by natural ventilation but can also
be designed with forced air. Free circulation of air must be main-
tained to provide satisfactory heat transfer.
Steel structures such as I beams, channels, plates and other me-
tallic, members either exposed or hidden should also be kept quite
away from the reactor and also they should not form closed circuit.
Winding conductor consists of one or more cables, depending upon
the current capacity, wound in radial layers. Cross-overs between
layers are made of minimize the eddy losses. Aluminium conductor
is normally used. Fibre glass insulation, the density of which plays
an important part in the overall rigid construction of reactor, is
adopted normally in the construction.
Porcelain insulators are used for achieving sufficient electrical
clearance to ground.
It is desirable that some safety enclosure which protects the
personnel and equipment should be put around the reactor. The 3
phase unit can be made up of 3 single phase coils either stacked
or arranged in one and the same plane. It is non-inflammable and
owing to its design, this equipment requires little supervision during
service. Maintenance is limited to clearing only.
19.2 Shunt Reactor
It extra high voltage networks, capacitive generation often creates

problem during operation at low loads and during switching opera-
tions. The severity of such problems increases with higher system
voltage, with extension of network and with longer line sections. In
most of the cases shunt reactors often represent an economically
and technically sound means of compensating part of the capacitive
generation.
1
2
5
4
1. Conductor
2. Conductor insulation
3. Insulator
4. Concrete supports and base
5. Provision for lifting
6. Terminal
Figure 19.10 Dry type current limiting reactor.
19.2.1 Types
The shunt reactor can be classified as follows:
(a) From the Service Point of View

— Normally in service
— Switched in, switched off type
If reactor is kept normally in service, during peak load time the
excitation of generator can be increased and consequently the stabil-
ity of the system. Under service fault conditions the reactor may be
switched off as a further measure to improve the margins of stabil-
ity. This disconnection momentarily increases the voltage of the net-
work. The use of shunt reactors in normal services may also result
in poor voltage levels in the underlying systems and increase system
losses. Depending upon loading condition, this may also necessitate
installation of static compensator banks for operation during peak
load condition. Switched-in reactor is normally used when system
Reactors 475
voltages are required to be controlled. The amount of reactance to be

switched in at a time must be limited to suitable value otherwise it
may result in objectionable voltage fluctuations in the entire system.
For switching in and switching off the shunt reactor, protective
gears should be more reliable. The switched in, switched out type of
reactor is sometimes connected to a low voltage tertiary winding of a
large transformer. The tertiary connected low voltage reactor is
obviously cheaper. But, it causes extra losses in the transformer. Its
rated voltage must be carefully selected because of the large voltage
drop in the series reactance of the transformer between high voltage
and tertiary winding.
(b) From the Construction Point of View

— Coreless (magnetically shielded)
— Gapped core
(Refer Clause No. 19.2.3)
19.2.2 General Consideration/Characteristics
Following characteristics are required to be considered:
(a) Highest System Voltage and Linearity

Normally the rated voltage for a reactor is defined on the highest
system voltage unless otherwise stated. Most of the specifications
require that the impedance of the reactor shall be linear to a certain
level above rated voltage.
The magnetization curve of the reactor is a relation between crest
values of current and flux. In order to avoid harmonic current
generation under system over voltage conditions or the risk of
non-linear ferroresonance of heavy inrush currents, shunt reactor
should have constant impedance up to about 1.3 times rated voltage.
Sometimes, constant impedance up to 1.5 times is specified. Further-
more, the impedance should be accurately balanced between phases
of 3 phase reactor. The acceptable tolerance on the current of any
single reactor at rated voltage is, according to IEC Pub. 289, ±5 per
cent of the rated current. Any one phase may deviate from the mean
of all three phases by ±2 per cent.
(b) Losses
The total losses in a shunt reactor from economical reasons should
be as low as possible. Low loss design with natural cooling generally
works out to be more economical. The tan delta figure of large

reactors is in the range of 0.30 per cent down to 0.15 per cent at the
highest capitilisation.
(c) Zero Phase Sequence Impedance

The requirement of zero phase sequence impedance is dependent
upon the system conditions. If this is 90 to 100% of positive sequence
impedance, coreless design with magnetically shielded or 5 limbed
gapped core reactor is most suitable. If this value is not very much
important or a value of about 50 to 60% of positive sequence imped-
ance is permissible, the 3 limbed gapped core reactor may be used.
(d) Single-phase or Three-phase Units

Large rating reactor sometimes poses problems of testing. To
overcome the above problem, banks of 3 single phase units are most
suitable. Further, the banks of three single phase units have the
following advantages:
— Zero phase sequence impedance is equal to positive sequence
impedance.
— Spare unit requirement will be only one single phase reactor
instead of a 3 phase reactor.
There phase reactors are manufactured for system voltages of 400
kV and below. It is technically feasible to extend the range up to 500
and 800 kV, but this has not been done yet. The reactors in 800 kV
class are usually banks of single phase units installed with a spare
reactor available in the station.
19.2.3 Construction
(a) Core-less
In a core-less reactor there is no magnetic material inside the coil
and the dimensions of the winding are identical to those of the air
gap volume, which imposes restrictions on the geometrical propor-
tions. The winding must be made wide and flat, so that the magnetic
flux path length is limited. The flux density in practice is much lower
than in the gapped core design and the total volume is larger for a
given reactive power rating. Flux with full intensity penetrates
the inner turns of the winding and gives rise to relatively high eddy
current losses. The flux density then sinks gradually towards the
Reactors 477
periphery of the coil, where it approaches zero. The average circum-

ference of the coil is appreciably larger than for the gapped core
reactor and the winding losses tend to be higher.
Core-less designs have a variety of alternative constructional
arrangements.
In USA, there has been frequent use of open, dry type coils
(Previously of concrete type). These are applied at low system
voltages for the compensation of cable systems. The design is
practically the same as for a current limiting reactor but it is larger
in dimensions.
Another design consists of a cylindrical coil with vertical axis,
placed in a cylindrical oil tank. The tank is lined with strips of core
steel which serve as magnetic shielding for the returning flux out-
side the coil. The whole coil volume is “air gap”.
Two cylindrical coils placed side by side between an upper and a
lower yoke of core steel form a single phase reactor where the flux
goes upward in one coil and down in the other. The coil volumes are
air gap. Three coils side by side in the same manner make a 3 phase
reactor. A compact non-circular pancake stack, surrounded by a rigid
frame of coresteel forms a type of reactor which is often referred to
as “shell form”, because of its similarity in shape and coil design with
a shell form transformer. The principle of shell type reactors gives
the possibility of arranging each phase in its own magnetic box.
Another type of construction without yokes, is a triangular confor-
mation of coils with the axis of each coil at the corners of a triangle
(Fig. 19.11). These is no yoke. The top and bottom of the tank are
screened from the coil flux by plates of high conductivity material.
Magnetic shields consisting of core pockets of core steel laminations,
all around the tank, restrict the flux linkage with tank thus reducing
stray losses and heating of tank.
(b) Gapped Core Reactor

The achieve high impedance of reactor in a core type construction
gaps having suitable size are inserted in the magnetic circuit. The
required effective length of the magnetic field is mainly dependent
on the dimension of the distributed gaps in the iron core and is
largely independent of winding axial height. The magnetic field
of gapped core is controlled by means of gaps. Compared with mag-
netically shielded coreless reactors, gapped core reactors can be
operated at higher flux density, the selection of which is dependent
upon the requirement of linear characteristic. Core elements are
constructed by using CRGO laminations. Various types of assembly
4
5
3
6
1. Line lead 4. Neutral line lead

2. Non-magnetic shield 5. Non-magnetic shield cut
3. Reactor away to show magnetic shield
6. Magnetic shield packets
Figure 19.11 Sectional arrangement of coreless shunt reactor.
of laminations are possible. Conventional type (transformer core

section), involute disc type and radially arranged moulded types are
normally employed to make the core elements. In modern reactors,
moulded type radial core sections moulded in epoxy resin to prevent
movement between individual laminations are commonly used. The
radial laminations prevent fringing flux from entering flat surfaces
of core steel, which would result in eddy current overheating and hot
spots.
Stiff material having very high dimensional stability under
pressure and high temperature conditions are used, to form gaps
between core elements. The core segments are accurately stacked
and cemented/glued together to make a solid core limb column. Top
Reactors 479
Welding of the core sheet

at the periphery of the tube
Core lamination
Tube
After shaping Earthing

After shaping the core
steel strips are
replaced by glass tape
impregnated to resin
Figure 19.12 Involute disc type core.
Cooling channels
Figure 19.13 Conventional type core.

and bottom yokes are butt or mitre jointed with legs. The core
elements are tightened under high pressure by tie rods and clamping
beams which maintain the pressure on core elements.
19.2.4 Noise and Vibration Problem

The magnetic field creates pulsating forces across air gaps which
amount to tens of tons. The reactor cores should, therefore, be very
stiff to eliminate objectionable vibrations. The spacers supporting
Figure 19.14 Gapped core with radially laminated core segment.
the air gaps should be of stiff material (usually ceramic blocks or

stone spacers) with a high modulus of elasticity so that their dimen-
sional stability is maintained and the core does not become loose.
Besides the gapped legs, care is also taken in the design of core
frame. The fundamental mechanical resonance mode of the reactor
frame must be designed to fall away from twice the power frequency.
The vibrations of active part are transferred to the tank through
mechanical contact at the bottom and also through the oil. Some-
times, antivibration pads are used between active part and tank
bottom. The tank walls set up a complex pattern of vibrations, and
there are possibilities for resonance of individual panels between
stiffener beams.
A large vibration amplitude in the middle of a large panel of thin
sheet may be less objectionable than a small amplitude on thicker
material. The resulting dynamic stress in tank and cover should
be limited to less than 20 N/mm2. Maximum vibration (double
amplitude) on tank wall at rated voltage should not exceed 0.2 mm.
Customer specifications also frequently specify sound levels and the
values are selected in analogy with those specified for transformers
with same MVA rating and BIL. NEMA TR-1 recommended levels for
transformers are generally applicable.
In the early days of shunt reactor design the noise and vibration
problems caused great concern to the users and manufacturers.
Sound level proved to be intolerable, and tanks developed oil leaks or
even cracked open due to vibration fatigue along weld seams. In
Reactors 481
modern designs, both gapped core and coreless with magnetic

shields, these problems have been well sorted out. High degree of
dimensional accuracy is required while manufacturing the core to
obtain reactance and phase currents within permissible limits.
19.2.5 Calculation Data

A reactor absorbs magnetic field energy. The energy per unit vol-
ume, or energy density in the field is
B◊H B2
=
2 2mm 0
The energy density shall be low, in core steel with high permeability.
The energy storage is concentrated in volumes with m = 1, i.e. in air
gaps. In a gapped core reactor, this storage volume consists of gaps
which are introduced in the magnetic circuit and in a coreless or
“shell form” reactor, the whole volume inside the magnetic shield
represents the energy storage.
Assume for simplicity, homogeneous field in the total volume.
The peak value of energy storage in the reactor is then,
B� 2
W= (Gap volume)
2m 0
Where B� = peak value of flux density at current maximum.

In terms of circuit elements
LI� 2
W= = L I2
2
Where I = rms value of current
Rated reactive power
B� 2
Q = wLI2 = wW = w . [gap volume]
2m 0
The gapped core design is based on high flux density in a relatively
small gap volume.
Also, Field energy per unit volume
B2
P = = Magnetic pulling force
2m 0
19.3 Neutral Earthing Reactor
Modern systems normally operate with their neutral points

grounded. Neutral earthing reactor is one of the means to ground
the neutral. There is no idea in purposely increasing the grounding
reactance of the system beyond that required to keep currents
within non-destructive range, except of course for the special case of
ground fault neutralizers used in resonant grounding.
In the transmission systems which require single pole opening
and reclosing of the EHV lines from the consideration of transient
stability, successful single pole reclosing requires that extinction of
secondary arc and the deionization of arc path in faulty phase should
occur before reclosing is effected. In this case it may be necessary to
reduce secondary arc current and thus may call for compensation of
phase to phase and phase to ground capacitance of the line. In EHV
system this is achieved by installing a single phase shunt reactor
(called neutral earthing reactor) between neutral point of EHV
reactor and the earth. Studies carried out have indicated that for
400 kV lines, basic insulation level of 550 kVp may be adequate for
the line terminal of neutral reactor.
The neutral earthing reactor is classified under the category of
current. For this reactor, no rated continuous current is applicable,
unless otherwise specified. The reactor should withstand without
undue heating or excessive mechanical stresses when short time
current is carried for a specified duration which is usually 10 sec.
Measurement of loss applies only when a continuous current is
specified. Induced over voltage withstand test purported to be routine
test is to be carried out at twice the voltage occurring across the
winding at rated short time current. In lieu of this test, lightning
impulse test is conducted and the impulse is applied to the terminal
which is to be connected to transformer or shunt reactor neutral,
whereas the other terminal is earthed. The duration of the front
time in such a case is permitted up to 13 microseconds.
19.4 Tuning for Filter Reactors
Tuning reactors have been defined as those reactors in a.c. systems

which are connected with capacitors to tuned filter circuits with
resonance in the audio frequency range for reducing, blocking or
filtering harmonics or communication frequencies. These reactors
Reactors 483
can be connected either in parallel configuration in the system or in

series configuration and can be oil immersed or dry type. Usually
these reactors are designed with means for adjusting the inductance
value within a limited range by means of tappings or by movement of
core and/or coils. Tuning reactors have rated quantities viz., continu-
ous current, voltage, inductance, Q factor at tuning frequencies.
19.5 Arc Suppression Reactors
Arc suppression reactors are single phase reactors used to compen-

sate for the capacitive current occurring in the case of line-to-earth
faults in a system with an insulated neutral. They are connected
between the neutral of a power transformer or an earthing
transformer and earth in a three-phase system. Sometimes a low
voltage secondary winding is provided for connection of a loading
resistor or for measuring purpose. Arc suppression reactors are
usually specified with their inductance variable either in steps or
continuously over a specified range to permit tuning with the
network capacitance. Adjustment of reactance can be accomplished
in one of the following ways:
(a) by adding additional sections of winding in finite steps with an
off-load or on-load tapchanger.
(b) by adjusting the air gap of the magnetic circuit by mechanical
means.
Compared to a neutral grounding reactor, arc suppression
coils have closer tolerances of rated current and currents at other
adjustments. Further a higher duration of rated current is envisaged
and temperature these limits are specified for continuous, 2 h and
30 minutes durations.
19.6 Earthing Transformers (Neutral Couplers)
Earthing transformers are three phase transformers or reactors and

are used to provide an artificial neutral for earthing of a system.
Earthing transformers are usually connected in zigzag. Some-
times, a secondary (low voltage) winding having a continuous rated
power for station auxiliary supply is provided.
19.7 Standards on Reactors
Indian Standard 5553 ‘Reactors’ is the governing standard on reac-

tors. This is in line with IEC 60289 (1988) reactors. It comprise of 8
parts instead of 3 parts of 1970 as follows:
Part I : 1989 – General
Part II : 1990 – Shunt reactors
Part III : 1990 – Current limiting reactors and Neutral
earthing reactors
Part IV : 1989 – Damping reactors
Part V : 1989 – Tuning or Filter reactors
Part VI : 1990 – Earthing transformers (Neutral Couplers)
Part VII : 1990 – Arc suppression reactors
Part VIII : 1990 – Smoothing reactors.
Almost entire range of important reactors are now covered. Spe-
cial purpose reactors like high frequency line taps or reactors
mounted on rolling stock are, however, not covered.
First revision of IS : 2026 (Part I to IV) is a necessary adjunct to IS
: 5553. References have been made to this standard at several places
specially for temperature rises and ambient conditions, dielectric
requirements and testing methods.
19.7.1 Transformer Standards

All transformer standards wherever applicable are also applied to
reactors. A few to mention are:
(i) IS : 335–1993 : Specification for New Insulating Oil (Fourth
Revision)
(ii) IS : 1866–2000 : Code of Practice for Electrical Maintenance
and Supervision of Mineral Insulating Oil in Equipment
(Third Revision)
(iii) IS : 2099–1986 : Specification for Bushings for Alternating
voltages above 1000 volts: (Second Revision)
19.7.2 CEA Report

The CEA report of “The committee for standardization of the
parameters and specifications of major items of 400 kV substation
equipment and transmission line materials” has standardized the
Reactors 485
specifications of 420 kV shunt reactors. Presently, requirement of

50, 63 and 80 MVAR reactors in Indian systems have been envisaged
and, therefore, only these ratings have been standardized. 1300 kVp
lightning and 1050 kVp switching impulse withstand levels have
been specified for these reactors. The neutral terminal of these star
connected reactors is earthed directly when 3 pole opening and
reclosing of 400 kV transmission line is practised. In case of single
pole opening and autoreclosing of lines, a neutral reactor is used for
earthing the neutral terminal. In such a case a BIL of 550 kVp is
considered necessary for the neutral terminal of shunt reactor.
The CEA report specifies 1.5 linearity for the V/I characteristic.
X0/X1 ratio has been specified to be between 0.9 to 1.0. A maximum
value of vibrations has been fixed as 200 microns anywhere on the
tank surface and the resulting permissible stress is 2 kg/mm2.
C HAPTER 20
Traction Transformers
J.M. Malik
The traction transformers require special considerations for their

design due to limitation of space availability and problems due to
vibration. They have to be designed and manufactured to stringent
specifications, so as to withstand heavy stresses in this type of
applications.
20.1 Types of Traction Transformers
The following types of traction transformers are in use with the elec-
tric locomotives of the Railways of the various countries.
20.1.1 Tapchanging Transformers for Low-voltage Control
The voltage is adjusted on the secondary side by a tapchanger or

tapping contactors. These types of traction transformers are gener-
ally used under the coach of Electrical Multiple Units with diode
rectifiers. The schematic diagram for the arrangement is given in
Fig. 20.1.
20.1.2 Tapped Transformers for High Voltage Control
The variable voltage is produced on the HV side of an auto-trans-

former by a tapchanger and is conveyed to a main transformer with
a fixed transformation ratio. This variant is used in powerful locomo/
tives with diode rectifiers. The schematic diagram for the arrange-
ment is given in Fig. 20.2.
Traction Transformers 487
A1 A2
HV
LV LV Aux. 1 Aux. 2
1 2 3 4 5 6 7 8
Figure 20.1 Schematic diagram of A.C. Electrical Multiple Unit

transformer.
20.1.3 Rectifier Transformers for Pulse-controlled Convertors
Mains convertors are fed through several secondary windings with a

fixed transformation ratio, e.g. six-pulse, four-quadrant convertors.
The schematic diagram for the arrangement is given in Fig. 20.3.
Here a polygon equivalent circuit diagram is illustrated to show
how a transformer with four secondary winding systems may be con-
nected.
The traction transformers for 25 kV AC locomotives using on-load
tapchanger for voltage control and traction transformers for thyris-
tor controlled locomotives are described in this chapter in detail.
20.2 Special Considerations
As distinct from an ordinary power transformer, the traction trans-

former has to withstand frequent mechanical and electrical shocks
and cater for overloads under adverse conditions. The traction trans-
formers have to withstand mechanical shocks in line with IEC-310,
which lays down a value of 3g in the direction of motion, 2g in the
transverse direction and 1g in the vertical direction, where g is the
value of acceleration due to gravity:
The following operating conditions must be taken into account,
particularly in design:
— Fluctuation of the contact wire voltage.
— Different short-circuit capacities of supply systems.
— High closing rate due to bouncing by the pantograph.
Through tap changer
V1 4 NR VR f II f III
fI 5
1 2 3
1, 2,3. Limbs
4. Regulating auto-transformer with steps
5. Primary winding of the main transformer
6. Secondary winding 1 of the main transformer
7. Secondary winding 2 of the main transformer
8. Auxiliary winding
Figure 20.2 Basic circuit diagram.
— Special climate conditions, such as when operating in long

tunnels or tropical areas.
The transformer windings should also suitably designed to with-
stand electromagnetic forces due to flash-overs in the rectifiers/
traction motors or electrical stresses due to the voltage surges
travelling along the exposed 25 kV overhead conductor system.
Any weakness in the insulation system or defective layout of the
windings could be fatal to the transformer. The voltage variations
OS
Z5
US3
X6 X10 US4
Z3 Z4
X7 X9
X8
US3 US4 OS1 US2 US1 OS2 Z2
Z1
US1 US2
OS1-Inner HV winding
OS2-Outer HV winding
US1-4-Four secondaries
Z1-Z5, X6-X10-Impedances of equivalent circuit diagram
Figure 20.3 Schematic diagram of four-quadrant convertors.
are unusually large in traction duty and form an important consider-

ation in the design of equipment. Indian Railways specify the following
values for this purpose.
Normal for traction 22.5 to 27.5 kV
Minimum for traction 19 kV
Minimum for auxiliaries 17.5 kV
Limitation due to locomotive design and other related conditions
are as under.
— Low weight
— Location of fixing
— Grading (the minimum possible number of steps in tractive
effort) and the high switching rate of tapchangers
— Auxiliary winding for supplying power to auxiliary equipment
and train lighting, etc.
20.3 Design and Constructional Features
20.3.1 Basic Design Concept
The basic circuit diagram of Fig. 20.2 shows the individual windings
and their arrangement on the three cores. The main transformer
(limb 2 and 3) is fed with variable voltage by the regulating auto-
transformer (limb 1).
(a) Regulating Auto-transformer and Main Transformer

The entire winding of regulating auto-transformer is divided into
individual groups which are suitably interleaved. The initial 80%
steps are of the same magnitude and consist of one double section
each, while the last 20% steps are somewhat smaller in magnitude.
As the currents on the series and parallel connection of traction
motors change from tap to tap, the transformer is designed for
continuous operation at each step with a constant secondary current.
The same limb 1 carries auxiliary winding also.
Both limbs 2 and 3 of main transformer are identical except the
core width. The primary winding of each core consists of two groups
connected in series. The low voltage groups are symmetrically
positioned in relation to the primary groups. The two secondary
branches have the same short-circuit voltage and are closely coupled
to each other. To obtain the optimum conductor cross-section, each
low-voltage winding has a number of conductors in parallel.
(b) Flux Relationship in Core and Iron-loss
The flux fI in the limb 1 is proportional to the supply voltage and

independent of the position of the tapchanging switch. The fluxes fII
and fIII in the two other limbs depend on the winding ratio NR/NP
and are as under.
\ f II =
fI F NR + 1I
2 H NP K
\ fIII =
fI F NR - 1I
2 H NP K
where
VI = incoming supply voltage
VR = regulating output voltage
N1 = series + common winding turns
NR = regulating winding turns
NP = primary winding turns
So the number of turns on one hand and the core cross-section
on the other are selected so that the flux in all three cores is approxi-
mately equal on traction operation steps, which correspond to
primary voltage of 22.5 to 27.5 kV.
(c) Copper-losses in Relation to Tapchanging Switch Position

The copper-losses of the regulating transformer very widely with
the tapchanging switch position, while the losses of the main trans-
former remain constant, since the secondary current remains
constant.
With a constant secondary current, the main transformer imped-
ance voltage varies with step number. This means, that the voltage
is lowest at the highest step and increases when the steps are low-
ered. To ensure that the overall impedance voltage is as constant a
characteristic as possible, it is essential to ensure that the regulating
transformer impedance voltage to be added does not vary much.
20.3.2 Constructional Details and other Features

(a) Core and Coils
Like any other power transformer, the traction transformer core is
made from cold rolled grain oriented steel. The core has a three-
limbed construction as shown in Fig. 20.4 (Plate 7).
Due to height limitation in electric locomotives, the core is placed
horizontally. It also requires careful tightening of the core, end
frames and cores together, to form a rigid assembly; otherwise there
is a possibility of sagging of the core limbs. The windings used
in traction transformer are of pancake construction as shown in
Fig. 20.5 (Plate 7).
The regulating winding and auxiliary winding are placed on limb
1, and limbs 2 and 3 are provided with HV and LV windings. The
conductors used are of paper covered high conductivity strips. As the
number of tapping leads are large they are carefully insulated and
anchored to prevent movement under mechanical forces. The entire
core and coil assembly is suspended from the tank cover and held
in position by steel pins at the bottom to prevent shifting of the
transformer.
(b) Insulation
Bulk of the insulation is precompressed pressboard, which does not
chip or get crushed under shock loads. The entire coil assembly is
carefully shrunk under predetermined clamping pressure and con-
trolled heat. As a further safeguard against any minor shrinkages
in service, spring loaded pressure pads are employed which exert a
steady pressure on all the windings through a sturdy steel plate, as
shown in Fig. 20.5. As the space is at premium, the precompressed
inter-section insulation is machined to close tolerances to avoid coil

build-ups. Insulation angle rings and angle barriers are placed
between HV and LV windings as shown in Fig. 20.5. They are of a
moulded construction, which brings compactness and mechanical
strength to the coil assembly. The insulation blocks are carefully
dimensioned and positioned between the windings, so as to ensure
adequate strength to the coils and effective cooling. All bolts and
nuts are properly secured to prevent loosening in service. All leads
are suitably clamped to preclude any possibility of a short-circuit due
to vibrations. The arrangement of lead connections to bus bars for
LV terminations and leads for tapchanger are shown in Fig. 20.6
(Plate 8).
(c) Tank and Accessories

The mild steel tank, oil conservator and cooling pipe-work are fabri-
cated to close tolerances to facilitate their mounting in the limited
space in the locomotive. In their design, care is exercised to keep
their weight down to a bare minimum but at the same time ensuring
sufficient strength against mechanical shocks. The tank rim and
cover are carefully matched to ensure oil tightness under the worst
operating conditions encountered during acceleration or retardation
of the locomotive, which bring in additional oil pressure at the joints.
The cover suspended transformer is provided with special screws to
adjust the height of the assembly, so that in fully tightened position
it sits evenly on base pads.
(d) Fittings
To bring in the desired compactness, the coolers are always of the
forced-oil forced-air type. The rate of oil flow and position of inlet and
outlet of oil from the transformer are governed by considerations of
optimum cooling. The oil pump is of a glandless-type construction
and is placed near the oil outlet. The oil passes through a cooler
consisting of finned tubes, over which the air is forced by a blower.
The oil from the cooler enters the transformer at the bottom and
then rises to the top near the outlet. To facilitate servicing and
replacement of the major components in the cooling circuit, valves
are introduced at the appropriate places. Figure 20.7 (Plate 8)
presents a view of the transformer with complete fittings. In Fig. 20.7,
to the right is the tapchanger with the conservator at the top and in
the centre is the blower motor for forced air cooling.
The traction transformer has an in-built system of protection
for maximum safety in service. This is achieved by a number of
protective devices such as, oil and air flow indicators in the cooling
circuits. The separately mounted CT in the primary circuit ensures
protection against excessive overloads and short-circuits. The
primary side is protected from the voltage surges by means of arcing
horns placed on the roof of the locomotive. The auxiliary windings
and associated equipment which are subjected to sudden rises in
voltages are protected by surge condensers connected across the
winding. The breather which prevents the ingress of moisture is also
of a special construction. The breather has a special shaped nozzle at
the entrance to filter out the heavy particles.
The transformer is provided with a condenser bushing for connec-
tion to 25 kV end. All other HV, LV, auxiliary bushings are porcelain
type and are mounted on top of the tank cover as shown in Fig. 20.7.
20.4 Traction Transformer for Thyristor Controlled

Locomotives
20.4.1 Introduction
The thyristorized controlled locomotive has the following advan-
tages, due to elimination of tapchanger as well as of the silicon
diodes.
— Smooth acceleration and retardation by stepless control.
— Full regenerative braking capability.
— Better adhesion and high speed detection and correction of
wheel slip and wheel skid.
— Facility of automatic current and speed control.
— Absence of wearing parts and consequent saving in main-
tenance costs.
20.4.2 Main Feature

The transformer is provided with a number of traction windings for
connection to thyristor system, an auxiliary winding for supply to
auxiliary machines and excitation winding to feed the traction motors
fields during rheostatic braking. The transformer design is suitable
for taking external connection of traction windings in series or
series—parallel combination. The transformer is provided with two
primary windings on two-limbed core and connected in parallel with
corresponding secondaries suitably magnetically linked, in order to
have minimum mutual coupling between the windings. The sche-

matic diagram of the transformer is shown in Fig. 20.8.
Core
A1
HV
A2
a4
LV4 a3
a2
LV1 a1
LVs
LV2 a8
a7
a6
LV3
a5
a10
Auxiliary
a9
a12
Excitation
a11
Figure 20.8 Schematic wiring diagram of thyristor-controlled

traction transformer.
The construction of the transformer is generally same as that of

the tapchanger controlled transformer, except that no regulating
winding is provided and core is two-limbed construction. The
construction of the transformer is such as to keep its weight and
dimensions as low as possible. In the design and construction of this
traction transformer, care is taken to incorporate all the salient
design and constructional features of traction transformer with
on-load-tap-changers as described in Secs. 20.3 and 20.4.
REFERENCES
1. Ecknaver, E. Traction Transformers for the Indian Railways,
Bulletin Oerlikon, No. 361.
2. Gupta, D.P., R.K. Shukla and J.M. Malik, Salient Features of
Traction Transformer, Electrical India, June 19/3.
3. Bohli. W.U., H.M. Deng and W. Muller, Transformers and Smoothing
Reactors for AC Traction Vehicles, Brown Boveri Review, Publication
No. CH–B 0540 E.
C HAPTER 21
Rectifier Transformers
J.S. Sastry
Chemical plants, aluminium plants, etc., where electrolytic processes

are adopted, require electrical energy in the form of direct current.
In these industries, the direct current requirement is so large that
it is considered to be one of the basic inputs for the production of
caustic soda, aluminium, etc. For example, about 3600 kWh of energy
is required for the production of one tonne of caustic soda and about
17000 kWh for aluminium.
Basically, the different methods for converting ac into dc are
(a) Motor generator set
(b) Synchronous convertor
(c) Mercury arc-rectifiers
(d) Rectifier transformer along with rectifier units using silicon
diodes
The advantages of using a rectifier transformer along with
rectifier unit is that, it can directly be connected to the supply lines.
This being a static unit, the efficiency is higher and enjoys all the
advantages of static unit. The third generation of electrochemical
plants have capacities as high as 250 kA. The present trend calls for
single unit of rectiformers having current rating as high as 100 kA
and more. Design and manufacture of such units pose many difficult
problems and place several constraints for the designer, which
require special attention.
21.1 Comparison between Rectifier Transformer

and Power Transformer
Though outwardly a rectifier transformer may appear to be similar

to a power or distribution transformer, it actually differs in many
respects. A power transformer is used to step the voltage up or down

for the transmission of the power, whereas a rectifier transformer
which steps down the voltage, is used in conjunction with rectifier
unit, for conversion of ac power into dc.
The salient features which make a rectifier transformer unique
when compared with power transformers are given hereunder:
(a) The general expression for direct current Id in the resistance
load ‘R’ connected to a transformer with m-secondary phases
and phase connected to a diode, assuming negligible diode
drop and transformer impedance, can be written as
+p /m
m Em cos w t
Id =
2p z
-p /m
R
dw t
Em m p
= ◊ ◊ sin
R p m
where, Em is peak phase to neutral voltage of the trans-
former. This is because each diode conducts for an interval of
360/m degree or 2p/m radians per cycle. Similarly, the rms
current per diode is
1 +p /m E2 cos2 w t
Irms =
2p z-p/m
m
R2
dw t
=
Em 1 pLM p
+ sin ◊ cos
p OP
R 2p m N m m Q
and the rms value of the current in the load having m pulses*
per second is m times Irms. The ripple value, defined by the
expression
2
Ripple =
LM I OP
rms
-1
NI Q d
The ripple decreases rapidly with increase in the number of
pulses or phases and hence the distortion of primary current
* Since waves are gradually changing alternating currents or voltages,

each cycle consists of a positive and a negative half, and the frequency
refers to how many cycle are completed in each second. In contrast a
pulse is an abruptly changing voltage or current which may or may
not repeat itself.
waveform is less and this is important since harmonics intro-

duced in the primary windings may cause interference in ad-
jacent telephone lines. This also raises the dc output voltage
and increases the load current. A rectifier transformer may,
therefore, have more than one LV winding and sometimes it
may have as many as four LV windings, unlike an ordinary
power transformer. Typical computed values of ripple are
Table 21.1 Computed Values of Ripple
m 2 3 6 12
Ripple frequency
for 50 Hz 100 150 300 600
Ripple 0.47 0.17 0.04 0.014
(b) Transformers which supply rectifiers do not carry sinusoidal

current in their windings or may carry currents in various
coils only over a portion of a cycle. The harmonic component
of these distorted waves contribute to transformer heating
but produce no useful dc output. To provide for these harmonic
currents, a transformer rating in watts, greater than the
rated dc power output, is required. Hence the Volt—Ampere
rating of the primary and secondary windings for rectifier
transformers may be different.
(c) The rectifier units connected to the rectifier transformer will
supply dc current for electrolytic processes as stated earlier.
This direct current has to be constant at all voltages. Hence a
rectifier transformer is basically a constant current trans-
former and consequently the KVA rating changes with output
voltage.
(d) The LVs of a rectifier transformer carry very high current.
So, heavy bus-bars will have to be used for the connections
between the windings and the terminals. The size and the
disposition of bus-bars carrying these high currents need
special attention so as to reduce the stray losses and also the
reactive drop.
(e) Other associated equipments of rectifier transformer are also
housed in the same tank. These may be:
(i) A regulating transformer which works on auto-trans-

former principle for voltage control.
(ii) Interphase transformer, if two star-connected rectifier
groups are to be paralleled.
(iii) A transductor unit for smooth voltage control on the
secondary side.
21.2 Rectifier Circuits
The main parameters of the rectifier transformer, viz. KVA rating of

the primary and secondary windings, currents and voltages, number
of secondary phases, etc., depend on the rectification circuits
adopted, which in turn are based on the requirements of the direct
current, voltage, ripple value, etc. Since the electrical parameters of
a rectifier transformer and its associated equipment have a direct
bearing on the circuit adopted, it is necessary here to briefly touch
upon some of the most commonly used circuits.
21.2.1 Six-Pulse System
(a) Double star-connection with Interphase Transformer

This gives a symmetrical load to the three-phase system. The
current transfer from one phase to the other occurs as soon as the
potential of the succeeding phase rises above that of the preceding
phase. Hence, the phase with the higher positive (or negative)
potential (depending on the polarity of the diodes) carries the load
current and determines the direct voltages.
Two three-phase connections are frequently paralleled through
an interphase transformer in order to double the load current and
to obtain six-pulse performance. This requires 180° phase shifted
transformer secondaries, i.e. one secondary connected in star and
the other inverted star as shown in Fig. 21.1. Each secondary
winding conducts during one-sixth of a cycle. Reflected to the
primary side, the secondary currents in the two windings on each
limb are opposite and 180° displaced and they can perfectly be
balanced by symmetrical primary current, regardless of the primary
connection. Since the primary side carries the current pulses of both
the secondary stars, the primary windings are effectively used but
Secondary-2
IPT
Secondary-1 L
o
a
d
Id
Figure 21.1 Six-pulse double star connection with interphase transformer.
the secondary rating is high as each secondary winding conducts

only for one-sixth of a cycle.
(b) Bridge Connection

In the connection described in (a) above, the transformer secondary
windings are idle most of the time, resulting in a high transformer
rating compared with the converted power. These windings can be
put to use more efficiently, if they feed bridge connected rectifier
circuit. This is accomplished by combining two three-pulse commu-
tating groups as illustrated in Fig. 21.2. A load connected between
points a and b is supplied with the total of two voltages when viewed
as separate groups. Since the two commutating groups operate with
opposite polarity, their ripple voltages are displaced and the total
voltage shows six-pulse ripple.
L
o
a
Secondary d
Figure 21.2 Six-pulse bridge connection.

However, interphase transformer connection is competitive over

the three-phase bridge connection in certain voltage-current range,
despite the relatively high transformer rating.
21.2.2 12-Pulse System

As already discussed in Sec. 21.1 (a), the lower the ripple of the direct
voltage and the distortion of the alternating line currents, the higher
is the pulse number. This is of minor importance for small power
supplies where simplicity is the first requirement, but it is the lead-
ing factor when selecting the connection for a large rectifier installa-
tion which usually consists of several independently commutating
units that are phase-shifted in such a way that the total pulse num-
ber is 12 or even higher. There is considerable reduction in the input
power and also the ripple content, if the pulse number is increased
form 6 to 12, but not much more if it is further increased. One of the
reasons for occasionally employing a higher pulse number is to re-
duce interference with other electrical systems, e.g. communication
systems, caused by harmonics in the alternating current.
(a) Quadruple Star-connection with Interphase Transformer

In this type of connection shown in Fig. 21.3, phase-shift of 30° is
achieved by providing two transformers, viz. one with the star
primary and the other with delta primary, each transformer having
two star-connected secondary windings, the neutrals of which are
connected to their associated interphase transformers. This would
require two rectifier cubicles placed on their side of the transformer
tank. It is important to note that the two cubicles are of opposite
polarities, so that the two fluxes oppose each other and minimize dc
magnetization of the transformer tank. The phase-shift in the pri-
mary windings can also be achieved by using zig-zag (interconnected
star) windings with +15° (leading) phase-shift on one transformer
and –15° (lagging) one the other, making it a total of 30°.
(b) Bridge Connection

A transformer with a primary star and secondary delta winding or
vice versa, causes a phase-shift of 30°.
Hence two complementary six-pulse systems can be built if one is
given a star-star or a delta-delta transformer and the other a star-
delta or a delta-star connected transformer. Even one transformer is
sufficient, if it has two secondary windings, one connected in star and
Secondaries
Primaries
Figure 21.3 Quadruple-star connection with interphase transformer.
the other in delta, provided that the rectifier circuits are bridge
connected as shown in Fig. 21.4. The two six-pulse systems can
either be connected in series or in parallel.
21.2.3 Transformer Utility Factor

As already stated, the windings of the transformer which supply to
rectifier units, carry currents only over a portion of a cycle. The
Secondaries
+
Primaries
Figure 21.4 12-pulse bridge connection.
harmonic components of those distorted waves contribute to trans-

former heating only, without any additional output. To provide
for these harmonics, the transformer equivalent kVA rating is
increased. The utility factor of a rectifier transformer winding may
be defined as the ratio of the dc power output to the equivalent
volt-ampere carried. It is a function of waveform and the ratio of rms
to average voltage. Table 21.2 gives the performance data for the
circuits described in this chapter.
21.3 Design Features of Rectifier Transformers
Depending on the circuit adopted, the tank of the rectifier trans-

former also houses any or all of the associated equipment listed
below:
(a) Regulating transformer — one
(b) Interphase transformer — one or two
(c) Transductor — one or two sets
A typical electrical scheme for 12-pulse system is shown in Fig. 21.5
for describing the important features of the above equipment.
21.3.1 Regulating Transformer

The technology used for the electrolytic and smelting processes
requires variable voltage. The ratio of the minimum to maximum
Table 21.2 Performance Data for Different Circuits
No. of sec. Bridge connection Interphase Transformer

phases connection
(m) 6 12 6 12
Conduction angle 60 30 64 30
Erms/phase secondary Edo/2.70 Edo/2.34 Edo/1.17 Edo/1.17
Irms of secondary 0.577 Id 0.408 Id 0.289 Id 0.144 Id
Peq of primary 1.28 Pdo 1.01 Pdo 1.05 Pdo 1.05 Pdo
Peq of secondary 1.28 Pdo 1.05 Pdo 1.48 Pdo 1.48 Pdo
Ptotal 1.05 Pdo 1.01 Pdo 1.05 Pdo 1.01 Pdo
PUF 0.78 0.99 0.95 0.95
SUF 0.78 0.95 0.625 0.625
Ripple value 4% 1.4% 4% 1.4%
The various symbols used in Table 21.2 represent the following:

Edo = Average value of theoretical direct voltage at no-load.
Id = Average value of direct current in load circuit.
Pdo = Edo ¥ Id
Erms = RMS value per phase of the primary or secondary alternating voltage at
no-load.
Irms = RMS value of the alternating current.
Peq = Equivalent rms volt-ampere rating of the transformer based on the
supply frequency.
Ptotal = Total volt-ampere input power (rms value), neglecting transformer ex-
citation current.
PUF = Primary utility factor.
SUF = Secondary utility factor.
secondary voltage is sometimes as high as 1 : 20. The realization

of this voltage regulation demands special considerations. The low-
voltage winding of the rectifier transformer is never provided with
tappings due to the obvious technical reasons. Hence, the desired
secondary voltage regulation is brought about by tappings arranged
on the high voltage side. This can be done in two ways.
(a) By providing taps on the primary of the rectifier transformer.
This is not always desirable due to technical and economical
considerations even in case of a six-pulse circuit, when the
voltage regulation is high. For 12-pulse circuits where
primary phase snift is done, this would be more complicated
and uneconomical.
Regulating transformer
Rectifier transformer-1 Transductor-1
I
P
T
1
Rectifier transformer-2 Transductor-2

Rectifier Transformers
I
P
T
2
505
Figure 21.5 Electrical scheme for a 12-pulse rectifier transformer.

(b) Feeding the primary of the rectifier transformer through a

separate auto-regulating transformer, which is a widely
accepted practice. The rectifier transformer being a constant-
current transformer, the output current of the regulating
transformer also is constant. Hence the KVA of the regulating
transformer changes with output voltage.
The voltage variation in the output voltage of the regulating
transformer can be achieved with the help of a regulating or tapped
winding. The tapped winding, besides serving the purpose of chang-
ing the output voltage within the given range, also maintains the
required output voltage when there is any variation in the input
supply voltage. For this purpose, a few taps on the regulating wind-
ing are intended to cater for the input supply voltage variation.
The output voltage of the regulating transformer is decided on the
parameters of the available on-load tapchangers, viz. the maximum
continuous current, voltage per step, voltage across the range and
the impulse level. Electrical location of the tap windings is largely a
choice of the designer and depends on the suitability of the on-load
tapchangers. The tapping arrangement can be either linear or re-
versing or coarse-fine type. Thus, it is seen that the selection of the
on-load tapchanger plays a vital role in the design of a regulating
transformer. The design philosophy of the regulating transformer is
very much similar to that of a normal auto-transformer.
21.3.2 Rectifier Transformer

This is also called the main transformer, as the output terminals
of this transformer are connected to the rectifier units. This is
essentially a stepdown transformer, the secondary voltage of which
corresponds to the output dc voltage. The HV terminals of this
transformer are connected to the output terminals of the regulating
transformer, if any, or directly to the supply lines in the absence of a
regulating transformer.
(a) HV Winding
This HV windings of this transformer are usually disc type and are
connected in either star or delta or zig-zag, which is described below.
It was mentioned that the phase shift between the input and
output voltage vectors can be achieved by zig-zag connection. This
can be explained with the help of Fig. 21.6(a) and (b).
A B C
120° Short
coils
Vi
q Vo
Long
(a) coils
(b)
Figure 21.6 Zig-zag connection.
We know that the three phases in a three-phase supply are 120°

apart. Now referring to the winding connections in Fig. 21.6(b), in
which two coils—one long and the other short, located on two differ-
ent limbs are connected in series. From the corresponding vector
diagram shown in Fig. 21.6(a), it would be noticed that the output
vector Vo lags the input vector Vi or in other words, there is a phase
shift of – q° between the input and the output vectors. A positive
phase shift can be obtained by slightly modifying the winding connec-
tions. The phase shift angle q depends on the ratio of the turns in
the short and long coils, which can easily be calculated. Table 21.3
gives the voltages of short and long coils as a percentage of the input
voltage for different phase shift angles normally in use.
(b) LV Winding
The secondary voltage of the rectifier transformer is very low and of
the order of a few hundred volts. Obviously, the turns in the second-
ary winding can be as few as two or three. On the other hand, cur-
rent carried by these windings is very high. For this reason a special
winding called half and sections (H and S) is used. This resembles the
Table 21.3 Variation of Phase Shift Angle
q Short coil Long coil Incoming voltage

in degree voltage in % voltage in % in %
3.75 7.552 96.010 100
7.5 15.072 91.609 100
10 20.051 88.455 100
11.25 22.527 86.815 100
15 29.886 81.650 100
20 39.493 74.220 100
22.5 44.188 70.294 100
30 50.000 50.000 100
disc winding to the extent that each coil is wound in two discs
accommodating the total number of turns. A number of coils of this
type are connected in parallel by bus-bars for sharing the total
current. The advantage of this coil is that odd number of turns can be
accommodated in two discs of the coil without any loss of space,
which cannot be done with the normal disc-winding technique. But
the condition for this is that there should be even number of parallel
conductors.
As an example, a coil with three turns and two conductors in
parallel is shown schematically in Figs. 21.7 and 21.8.
4
Y X
X Y
5 3
6 2
7 1
8
Disc. 1 Disc. 2
(a) (b)
Figure 21.7 Discs of half and section.

1 2 3 4 XY 5 6 7 8 9
X
Y
X X
Y
X
Y
X
Y Y
X
Y
X Y
Figure 21.8 Development of half and section winding discs.
(c) Interleaving of LV Windings

The two groups of coils of both the secondaries are axially inter-
leaved as shown in Fig. 21.9. The interleaving is done to ensure that
the impedance between the two secondaries is minimum and the
impedance of the secondaries with respect to the primary windings is
same so that all the coils of both the secondaries share the current
equally. The ends of the bus-bars are then connected in star or delta
as the case may be.
(d) Coil Disposition

Since the secondary coils carry heavy currents and the coils are
connected in parallel by means of bus-bars, it is essential that the
secondary coils are placed outermost for ease of connection. Conse-
quently, the HV coil is placed concentrically over the core, over
which the LV coils are placed.
(e) Bus-bars
In addition to carrying LV currents, bus-bars are also used for
neutral formation, neutrals to interphase transformer connection
and for connecting the LV bus-bars to the transductors. The design
of the bus-bars and the arrangement has to be keeping the following
important considerations in view.
(i) Skin effect. A reasonably accurate determination of the current
rating of the conductors carrying direct current is possible from the
theoretical considerations, but with heavy alternating currents the
Secondary-1 Secondary-2
Line Neutral Line Neutral
Dise. No.
1
Secondary-1
2
1
2
Secondary-2
1
2
1
2
Secondary-1
1
2
1
2
Secondary-2
1
2
Secondary-1 1
2
Figure 21.9 Axial disposition of secondary windings.
current density is not uniform due to the skin effect and the apparent
ac resistance is always greater than that measured by dc methods. In
the case of flat bus bars, skin effect results in a concentration of
current at the edges of the bus-bars. With the larger sizes of the
conductors for a given cross-sectional area, skin effect in a thin flat
bar or strip is usually less than in a circular rod but greater than in a
tube. It is dependent, among other things on the ratio of the width to
the thickness of the bar and increases as the thickness of the bar
increases. The curves in Fig. 21.10 give approximate skin effect
ratios for flat bars of various cross-sectional areas and proportions.
The skin effect ratio is the ratio of the apparent ac resistance of a
conductor (R¢) to its dc resistance (R).
(ii) Proximity effect. In the foregoing consideration of the skin effect,
it has been assumed that the conductor is isolated and at such a
distance from the return conductor that the effect of the current in it
can be neglected. When conductors are close together, as is often the
8 a/b
1.4
16
8
b 16
cy ycle
cle
50 0 c
1.3 a
6
24
R¢
R
1.2 8
16
24
cle
cy
25
1.1
1.0
1 2 3
Figure 21.10 Skin effect ratios for flat bus-bars.
case, a further distortion of current density results from the interac-

tion of magnetic fields of the other conductors. The magnetic flux of
one conductor may produce emf by induction in any other conductor
sufficiently nearby. If two such conductors carry currents in opposite
directions, their electromagnetic fields are opposed to one another
and tend to force one another apart. There is, therefore, a decrease
of the flux linkages around the adjacent parts of the conductors and
an increase around the more remote parts, which results in a con-
centration of current in the adjacent parts where the opposing emf is
minimum. If the currents in conductors are in the same direction,
the action is reversed and they tend to crowd into the more remote
parts of the conductor. This is called, the proximity effect. In some
cases, however, proximity effect may tend to neutralize the skin
effect and produce a better distribution point, as in the case of strip
conductor arranged with their flat sides, towards one another. If
such conductors are arranged edgewise to one another, the proxim-
ity effect increases the skin effect.
(iii) Bus-bar reactance: An accurate knowledge of bus-bar reactance
is necessary, as this would directly add to the transformer reactance
for estimating the regulation drop under fully loaded conditions. The
bus-bar reactance is calculated in two parts. The first part consists of
“Go” and “Return” path, i.e. up to the point where the line and
neutral bus-bars (risers) run together. The second part is for single-
phase path, i.e. from the point where neutral is made to the point of
connection to the line terminals. In calculating this reactance, it is
assumed that in the case of laminated bus-bars connected in parallel,
provided the spacing between the individual bars is small, the
reactance may be taken to be the same as that of a single conductor
having the same overall dimensions as the group.
1. Go and return path. Referring to Fig. 21.11, let the values
of a, b and d be as shown. The ratios (b/a) and (d/a) are calcu-
lated and the corresponding Dm, geometric mean distance
between two paths is read off from the curve given in
Fig. 21.12. Then Ds, the geometric mean distance* of each
path from itself is calculated from Ds = 0.2235 ¥ (a + b).
Then associated inductance La is calculated by equation
La = 0.002 ¥ l ¥ log (DmDs) mH
where l is the length of “Go” and “Return” path. Hence, the
associated inductance La included both self and mutual
effects, i.e. La = L – M. From this inductance, reactance can
easily be calculated.
Figure 21.11 ‘Go’ and ‘return’ bus-bars.
2. Single-phase path. As before, the value of Ds is calculated.

Then inductance L is given by L = 0.002 ¥ l ¥
* Bus-bar sections are not filaments of negligible thickness, leaving no

doubt as to the distance between one bus-bar and another, in fact they
are large finite sections. The distance between two points may be
different from that between any other pair and hence the concept of
geometrical mean distance Dm.
b
1.0 0
0.9
0.8
0.7
0.6
0.5 0.3
0.4
0.3
0.2
0.5
0.6
0.1
0.09 0.7
0.08
0.07
0.06 0.8
Dm –1 0.05
d 0.04
0.03 0.9
0.02
0.01
0.009
0.008
0.007 1
0.006
0.005
0.004
0.003
0.002
b
0.001 a =1 0.9 0.8 0.7 0.6 0.5
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 d 2 3 4 5 6 7 8 9 10
0.9 1.0 a
Figure 21.12 Geometric mean distance.

FG log 2l - l + D IJ mH where l is the average length of the

s
H D s L K
single-phase path. From the above inductance, the reactive
drop for the single-phase path can again be calculated.
The sum of the above two paths gives the reactive drop
introduced by the bus-bars.
21.3.3 Interphase Transformer

A number of batteries can be paralleled only when their terminal
voltages are equal, as otherwise there will be internal circulating
current. Parallel operation of two rectifier systems is more
complicated as their direct voltages are fluctuating. Such systems
can be connected in parallel without any circulating current, if their
direct voltages are equal at any instant, i.e. if the average values are
equal and ripple voltages coincide. Usually, however, it is desired
that the ripple voltage instead of coinciding, be so displaced that
the combination results in a system with higher pulse number. The
parallel connection, then, must be made in such a manner that it
does not effect the operation of the individual group. Through an
interphase transformer, two rectifier systems with displaced ripple
voltages are paralleled. It absorbs, at any instant, the difference
between the direct voltages of the individual systems and must be
designed for the time integral of this voltage.
Figure 21.13 shows the direct currents through an interphase
transformer. This transformer absorbs at any instance, the voltage
difference between the individual groups and thus maintains
independent operation of these groups.
With respect to the voltage difference to be absorbed, the two
windings of the interphase transformer of Fig. 21.13 are in series
connection. Thus, the voltage difference can be balanced by the emf
induced in these windings—just as in normal transformer. The volt-
age impressed on the primary windings is balanced by the induced
emf. However, inducing the balancing emf needs a changing mag-
netic flux and exciting ampere-turns, which is the difference of the
direct currents to be combined (since these currents pass in opposite
direction through the window of the interphase transformer).
If these currents are well balanced, the core will not be driven
into saturation by the dc ampere-turns, even without an air gap. The
time integral of the voltage to be absorbed by the interphase trans-
System-A Interphase transformer
System-B
–
+
Figure 21.13 Paths of direct currents in interphase transformer.
former is a function of the dc voltage, of the operating conditions

(angles of phase control and commutation) and of pulse number. For
regular transformer, operating at 50 Hz, the maximum flux density
is chosen in the vicinity of 1.6 tesla. For an interphase transformer,
a lower value, say 0.9 tesla, is taken, since the magnetic flux is alter-
nating with three times the supply frequency, if the two three-pulse
systems are combined.
Design Features
An interphase transformer can be either a wound type or a bar-
mounted type depending on the voltage to be absorbed and the direct
current rating.
(a) Wound type interphase transformer. The wound type interphase
transformer is like a single-phase transformer with windings on the
two legs of the core.
The total number of turns in each windings is divided equally and
accommodated on the two limbs. The connection diagram is as
shown in Fig. 21.14. The terminals N1 and N2 are connected to the
respective neutrals of the two secondaries of the rectifier trans-
former and N forms the negative terminal of the dc supply. Since the
windings carry large currents and the number of turns if few, H and
S coils are used.
(b) Bar-mounted type of interphase transformer. In this type of
interphase transformer, the bus-bars from the neutrals pass through
rectangular cores forming a single turn winding. The arrangement
is shown in Fig. 21.15. As before N1 and N2 are connected to the two
neutrals and N to the dc negative terminal.
N1 I N N2
I/2 I/2
I/2 I/2
Figure 21.14 Schematic diagram of wound type interphase transformer.
N
I/2 I/2 I I/2 I/2
N1 N2
Figure 21.15 Schematic diagram of bar-mounted type interphase

transformer.
21.4 Transductors
A transductor is a device consisting of one or more ferromagnetic

cores with windings, by means of which an alternating voltage or
current can be varied by a direct voltage or current utilizing the
saturation phenomenon of the core material. Coarse control of the
rectifier output voltage is achieved by changing the tapchanger posi-
tion, i.e. increase or decrease of the output voltage can be achieved
in steps by the tapchanger. The function of the transductor in a rec-
tifier system is to smoothly control the output voltage between any
two tapchanger positions. The transductor is usually designed to
cover a range of two steps of voltage of the tapchanger, so that the
frequency of operation of the tapchanger is reduced.
It is well known that the magnetization produced in an iron core
by passing current through a coil wound round it, is not constant, but
depends on the magnitude of the current. This property can be used
to control the power in a circuit in a convenient and flexible manner.
With increase in current, the magnetization increase rapidly at
first and is roughly proportional to the current but soon the magne-
tization curve begins to flatten out. Increasing the current still pro-
duces increased magnetization but not in the same proportion and
finally a point is reached when further increase in the current pro-
duces no change in the magnetic field intensity.
In this condition, the iron is said to be saturated. If an alternating
voltage is applied across the coil, the current which flows, neglecting
resistance, is determined by the inductive reactance, i.e. the product
of the angular frequency of the current and the inductance of the
oil. The inductance of the coil is proportional to the degree of magne-
tization. In other words, current which will flow, is dependent on the
state of iron core. Suppose, therefore, we place two coils on the
core as shown in Fig. 21.16. One is a control winding connected to a
source of dc which serves to determine the state of the iron core.
But the other is connected to a source of ac supply. The circuit is
completed through a load.
The load current will be determined by the combined effect of the
coil L and the load R. If the inductance of the coil is small, the whole
of the applied voltage appears across the load. On the other hand, if
the inductance is high, most of the applied voltage will be expended
in over-coming the impedance, so that the current through the load
becomes very small. But we have seen that the inductance of the coil
L depends on the state of iron which can be controlled by current in
AC
supply
DC supply
Load
Figure 21.16 Transductor schematic diagram.
the dc winding. Hence, by suitable adjustment of the direct control

current, the power in the load can be regulated within wide limits.
In normal practice, there will be two dc windings–called bias
winding and control winding. The current through the bias winding
is preset at a particular level and the current through the control
winding is smoothly varied.
The usual form of transductor core is toroidal, through which
the secondary bus-bars of the rectifier transformer are threaded as
shown in Fig. 21.17.
The control and bias windings are in the form of copper rod, also
threaded through the ring core. The ring cores are made of CRGO
steel. A number of such cores, depending on the voltage to be
absorbed or released (or simply the range of the transductors) are
clamped between two permawood boards by means of insulated
bolts. There will be as many stacks of the transductor as the number
of the secondary phases. Figure 21.18 (Plate 8) shows the photograph
of assembly of transductors mounted on the top clamp plates of the
rectifier transformer Fig. 21.18.
21.5 Constructional Features of Rectifier Transformers
21.5.1 Types of Construction

Use of silicon rectifiers along with rectifier transformers as a heavy
dc power source for the electrolytic industry has come to be wide
spread. The relative placement of these two units is of considerable
importance from the economics and operational point of view. This is
End board
1
Ring core
2
7
End board
Bias rod
Control rod
LV bars
TIE rods
Figure 21.17 Transductor assembly.
due to the fact that enormous conductors carrying heavy currents

are necessary between rectifier units and the transformer. The
length of the connectors has the following implications:
(a) The cost of the conductors and their installation account for
a considerable part of the total construction cost of the
equipment.
(b) The current carried by this conductor is of rectangular
waveform and consequently eddy loss is inevitable. This,
together with high current lowers the overall efficiency of the

equipment.
(c) Inductive voltage drop due to the heavy currents lowers the
power factor, and also the regulation becomes poorer.
(d) Strong magnetic field around the conductor carrying currents
causes inductive heating. For this reason, the conductors
support frame must be made of a non-magnetic insulators.
Hence great attention has to be paid to the arrangement of the
rectifier cubicle and the transformer, so as to make the connection
between their terminals as short as possible. This has been success-
fully realized in the Rectiformer construction, in which the trans-
former and the rectifier cubicles are installed as near as possible
longitudinally. Figure 21.19 shows the schematic arrangement of
the rectiformer construction for a 12-pulse system. The secondary
terminals of the two rectifier transformers are brought out on either
side through the longer side-walls of the tank for ease of connection
to the reactifier cubicles.
Rectifier cubicle
Transformer secondary
terminals
Transformer
Rectifier cubicle
Figure 21.19 Rectiformer construction.
In the other type of construction, which is a conventional one, the

transformer and the rectifier can be installed at different places and
connected together by bus-bars. This would facilitate the installation
of the transformer in the outdoors switch-yard outside the building
which houses the rectifiers and the electrolytic cells.
21.5.2 Layout
It has been stated earlier that all the associated equipment of
the rectifier transformer, like regulating transformers, interphase
transformers and transductors are placed in the tank for obvious
advantages like smaller space requirement and higher operational
reliability. In addition to this the other advantages are:
(a) By placing the regulating transformer in the same tank not
only considerable space is saved, but it also reduces the risk
of the short-circuit on outgoing terminals of the regulating
transformer. It obviates the need for expensive short-circuit
protection gear owing to their low impedance voltage,
specially when the ratio is small (auto-transformers are not
short-circuit-proof on their own).
(b) Materials like steel, cables, oil, copper bus-bars, etc., can be
saved by this arrangement. This is particularly the case
with the interphase transformer and transductor, since they
accommodate heavy copper bus-bars. This also results in
lower bus-bar reactance, losses, etc. A typical layout of all the
equipment mentioned is shown in Fig. 21.20, which shows the
internal arrangement of rectifier transformer tank for 12-
pulse, with the regulating transformer placed perpendicular
to the main transformers. The interphase transformers are
placed over the top of the main transformers and the
transductors in a separate pocket in front on the main
transformers.
21.6 Tank Design
The tank dimensions are decided keeping the magnetic clearances in

view. The magnetic clearances are very important because the bus-
bars carrying heavy currents in the proximity of the structural steel
would cause additional stray losses due to the flux linkages. This
phenomenon is also called the proximity effect. The additional losses
caused by placing a magnetic material close to bus-bar carrying cur-
rent arise from both hysteresis loss and eddy current effects within
the plate. Hysteresis losses are large if the flux density within the
plate is high. The material having high permeability and very high
Transductor
transformer
Regulating
Main transformer
Transductor-1
Main transformer-1
Interphase transformer-1
transformer
Regulating
Interphase transformer-2
Main transformer-2
Transductor-2
Figure 21.20 Internal layout of rectifier transformer.

resistivity would increase hysteresis loss, because the flux developed

by bus-bar current would concentrate within the low reluctance
plates, and because the action of eddy current to counteract the inci-
dent flux would be comparatively small in a high resistance material.
Eddy current losses depend on the magnetic field strength at the
plate and also upon the resistance of the paths available for flow
within the plate.
These additional stray-losses can be controlled by the following
methods:
(a) By interleaving the bus-bars, so that only the resultant flux
links with the tank wall.
(b) By placing the bus-bars in such a way that their width is
always perpendicular to the tank wall. It can easily be shown
that the loss in the steel plate is far less when the flat surface
of the bus-bar is at right angles to the steel plate than when
it is parallel.
(c) By increasing the distance between the bus-bars and the tank
wall.
(d) The magnetic plate can be shielded with a sheet of good
conducting material, such as aluminium, placed so that the
magnetic field acts to build up counteracting circulating
currents within the conducting sheet. These current reduce
the magnetic field strength near the steel plate considerably.
(e) Non-magnetic steel inserts can be introduced judiciously, or
the magnetic plate can entirely be replaced by non-magnetic
steel plate. Non-magnetic steel has low permeability and high
resistivity resulting in loss reduction. The strategic locations
where antimagnetic stainless steel plate is provided are
where the secondary or interphase transformer terminal
bushings are located.
21.7 Testing
The rectifier transformers are designed and manufactured conform-

ing to national or international standards. Due to the complexities of
the circuitry of the transformer and its associated equipment, special
methods will have to be devised for conducting some of the tests.
Only such tests are described hereunder:
21.7.1 Measurement of Winding Resistance

Since the output leads of the regulating transformer are connected
to the hv of the rectifier transformer inside the tank, the rectifier
transformer hv winding resistance will add in series to that of the
regulating transformers winding during the winding resistance
measurement. For the accurate measurement of the resistance of
the individual windings, for the computation of load losses, and
further for the measurement of hot winding resistance at the end
of temperature rise test, these two transformers will have to be
temporarily isolated. One of the methods is to isolate at the diverter
switch if these contacts are accessible externally.
21.7.2 Impedance and Load Loss Measurement

These tests are simultaneously conducted on normal power trans-
formers by short-circuit method. However, in case of rectifier trans-
former where the primary and secondary ratings are different (e.g.
the rectifier transformer having double star-connection with inter-
phase transformer) and since the impedance is declared on the basis
of the primary rating and the load losses are declared on the basis of
the load losses in the individual windings at rated currents, this test
is conducted separately as below:
(a) For the impedance voltage measurement, all the secondary
terminals are shorted and the voltage required for the full
rated primary current to flow is applied to the primary.
(b) The load-loss is measured first with one secondary shorted
and the other secondary open. The test is repeated with the
other secondary shorted and the first one open. During this
test, the current in the primary is adjusted to its full rated
current. The average of the two measured loss figures gives
the total load-loss in the transformer.
Publication 146/1973 –“Semiconductor convertors” (also IS :
4540–1968–Specification for mono-crystalline semi-conductor recti-
fier assemblies and equipment), stipulates clearly the test procedure
to be followed for load-loss measurement for various circuits that are
in practice.
REFERENCES
1. Copper for Bus-bars, Copper development association, U.K.
2. Reyner, J.H., Magnetic Amplifiers, Rockliff, 2nd ed., 1956.
3. Ryder, J.D., Engineering Electronics with industrial application and
control, McGraw-Hill, Kogakusha Co., New York, 1957.
4. Schaefer, Johannes, Rectifier Circuits—Theory and Design. Wiley &
Sons, New York, 1965.
5. Brown Bovery Review—March/April 1961.
C HAPTER 22
Convertor Transformers
for HVDC Systems
I.C. Tayal
C.M. Sharma
S.C. Bhageria
The transformers used in HVDC systems usually named as the

convertor transformers have different design requirements, due to
superimposed DC voltages and currents. Convertor transformers for
12 pulse rectification have three windings, one star connected AC
winding and two valve windings–one connected in delta and the
other in star. The AC winding is connected to AC system, whereas
the valve windings are connected to convertors which, in turn, are
connected in series to build up the required level of DC voltage.
Figure 22.1 shows simplified connection diagram for convertor trans-
formers connected in a typical HVDC system. Due to typical location
it has to withstand various abnormal and critical conditions which
demand special design and manufacturing features.
Insulation system of a convertor transformer has to withstand not

only AC voltages and short time over voltages, but also superim-
posed DC voltages on AC voltages and polarity reversal.
It is noteworthy that while the voltage stress distribution under
AC conditions depends on inverse ratio of the dielectric constants
of the insulating materials, the DC voltage stress distribution is
dependent on the direct ratio of the resistivities.
Convertor Transformers for HVDC Systems 527
DC pole
AC busbar
Convertor Quadri valve

transformers 12 pulse
DC valve
To ground electrode
Figure 22.1 Convertor transformer connected in a typical

HVDC system.
Major insulating materials used in oil barrier insulation system

of convertor transformers are oil, paper and precompressed press-
board. The ratio of the dielectric constants for these materials does
not vary more than 1 : 2 and is practically independent of external
factors. Therefore, the stress distribution under AC is more uniform.
Figure 22.2 shows typical AC field distribution in the end insulation.
The stress distribution under polarity reversal is capacitive in
the beginning and oil is stressed more than the solid. It will then
successively change over to resistive distribution. The ratio of the
resistivities of the solid insulating material and oil varies by 2 to 3
power of ten depending on many external factors. Therefore, the
solid insulation is stressed more under the DC voltage. Figure 22.3
shows DC field distribution in the end insulation for the ratio of
resistivities 1 : 100.
Partial discharges under DC stresses appear in the form of
sporadic pulses at random intervals and manifest itself in different
ways:
0%
100%
Figure 22.2 AC field distribution in the end insulation.
100 %
Figure 22.3 DC field distribution in the end insulation.
(i) Rapid changes in voltage in oil gap may cause discharges

under certain conditions. Generally in well designed and
manufactured insulation systems, low energy level dis-
charges take place which have practically very little effect
on insulation.
(ii) Discharges in cellulose insulation generally take place due to

imperfections in the insulation. Such discharges slowly
disappear as the voltage across the solid insulation decreases.
Proper design approach, selection of appropriate materials
and careful manufacture ensure elimination or reduction
of such discharges as otherwise these tend to reduce the
impulse withstand strength.
(iii) Discharges at the oil cellulose interface are also controlled
through adequate design and manufacturing practices.
PD measurement is made during long time DC voltage test and
DC polarity reversal test to estimate the extent of these discharges.
During the last 10 minutes of the long time DC voltage test, not
more than 10 pulses above 2000 pC are allowed and not more than 30
pulses are allowed during the last 29 minutes of the DC polarity
reversal test.
The different nature of voltage distribution under AC and DC re-
quires different arrangements of the oil-barrier insulation system
than is usually necessary for conventional power transformers. Con-
siderations must be taken for this in designing the windings and
insulation system.
22.2 Higher Harmonic Currents
The harmonic content in a convertor transformer is much higher

than a conventional AC transformer. There are additional load losses
due to the presence of harmonic currents in valve windings. High
leakage flux is produced by these harmonics. The stray flux due to
these harmonics causes additional stray and eddy losses not only in
winding conductors but also in transformer tank and steel structure.
These losses may cause local hot spots. Avoidance of such occur-
rence requires proper selection of winding conductors in critical
fields/zones, provision of suitable magnetic shunts, directed oil flow
for effective cooling along with additional cooling equipment.
22.3 DC Magnetization
Inaccuracies in valve firing (i.e., firing asymmetry) give rise to DC

magnetizing component. Although, with modern firing control, the
DC excitation component is much lower, yet careful selection of suit-

able flux density is imperative to avoid excessive losses in the core
due to presence of DC current.
22.4 DC Bushings
The behaviour of an arc under DC and AC conditions is different and,

thus, the bushing for DC terminal has to be different. The DC with-
stand voltage of contaminated insulator is 20% to 30% to that of AC.
For satisfactory performance, bushings with higher creepage is re-
quired to avoid frequent flashovers. Creepage distance as high as 40
mm/kV is generally specified.
22.5 On Load Tapchanger
Due to the large range of voltage control requirement at both conver-

tor and invertor ends, the tapping range of convertor transformers is
generally much wider than conventional AC transformers. Mixed
voltage control is also required to meet the voltage requirement.
Such requirements call for special tapchanger with higher tapping
range and with appropriate make and break capacity.
22.6 Influence of Impedance Variation
Any variation in the specified impedance for a convertor transformer

influences cost of associated equipment in a convertor station.
Therefore, the permissible tolerance on impedance of convertor
transformer is much lower than that of a conventional transformer.
Generally 6% tolerance on the specified impedance and 2–3% varia-
tion in measured impedances of different units against a particular
design are allowed while in case of conventional transformers
acceptable tolerance on the specified impedance is ±10%.
Closer tolerance on the difference in impedance of AC winding
to star connected and delta connected valve windings of convertor
transformer, reduces the distortion in the DC voltage wave form.
Similarly, lesser variation in impedance between two convertor
transformers in a convertor station reduces non-characteristic
harmonics, thereby reducing cost of AC filters. Closer tolerances

on tested impedances of various units permit optimum design of
convertor station.
22.7 Connections
22.7.1 3-Phase, 2 Winding (Fig. 22.4)

In this case two designs and one unit of each design is needed for a 12
pulse convertor. One spare unit of each design is recommended.
Figure 22.4 Connection diagram of 3-phase, 2 winding.
22.7.2 1 Phase, 2 Winding (Fig. 22.5)
In this case 2 designs and 3 units of each design are needed for 12
pulse rectification. One spare unit of each design is recommended.
22.7.3 1 Phase, 3 Winding (Fig. 22.6)

In this case one design with two wound limbs, one limb Y-Y connected
(in three phase bank), the other Y-D connected (in three phase bank)
are required. Three units are needed for 12 pulse convertor. One
spare unit is recommended.
22.7.4 Extended Delta Connection (Fig. 22.7)

One 3-phase unit or three 1-phase units for a 12 pulse convertor are
required. One spare unit is recommended.
Figure 22.5 Connection diagram for 1 phase, 2 winding.
Figure 22.6 Connection diagram of 1 phase, 3 winding.
Figure 22.7 Diagram of extended delta connection.

22.7.5 3-Phase, 3 Winding (Fig. 22.8)
One 3-phase unit for 12 pulse convertor is required. One spare unit is
recommended.
Figure 22.8 Connection diagram 3-phase, 3 winding.
22.8 Specifications
The specification of a convertor transformer, due to the presence of

DC voltage and current, is different from that of conventional trans-
former.
The salient features of specification of convertor transformers for
National HVDC experimental line project and Rihand-Delhi NTPC
Commercial line are given in Tables 22.1 and 22.2 respectively.
Figure 22.9 (Plate 8) shows a photograph of the BHEL make 40.5
MVA convertor transformer for the National HVDC project.
22.9 Manufacturing Features
In order to meet onerous demands of a convertor transformer,

manufacturing practices have to be adequate to ensure the follow-
ing:
— Proper selection of materials and sources of materials,
particularly insulation items, is essential to ensure discharge-
free insulation structure, both solid and liquid in AC, DC and
polarity reversal conditions.
— Close dimensional tolerances in all windings of transformers
are essential to achieve desired impedance and also to ensure
minimum deviation between calculated and tested impedance.
Table 22.1 Specification of Convertor Transformer for National HVDC
1. kVA Rating : 40,500 kVA, Single phase

220 kV
2. Rated Voltage HV :
3
86.6 kV
LV :
3
3. HV-LV impedance at : 19.3%
principal tap
4. Connections : Star/Star
5. Tappings : 26 steps, On load
Range : –10% to + 22.5%
Numbers of steps : 13 steps, reversing arrangement
and variation for +/–10% HV voltage variation
and at +10% HV voltage 0 to –10%
LV voltage variation
6. Type of cooling : OFAF/ONAF/ONAN
7. Terminals HV : AC Bushing
LV : DC Bushing
8. Insulation levels
8.1 Impulse HV : 950 kVp Lightning
: 750 kVp Switching
LV : 450 kVp Lightning
450 kVp Switching
8.2 Induced voltage : HV 395 kV, LV 185 kV
(fully insulated)
8.3 Long duration DC voltage
8.3 test associated with
8.3 partial discharge : 166 kV for one hour
8.3 Short duration DC
8.3 voltage test : 220 kV for 2 minutes
8.3 Direct voltage polarity
8.3 reversal test associated
8.3 with partial discharge : 88.34 kV
— When several units are required in a convertor station,

dimensional accuracies of coils have to be identical to ensure
that the difference between impedances of different units is
minimum. This calls for highly accurate moulds for windings,
utilisation of appropriate winding machines with provision for
Table 22.2 Specification of Convertor Transformers for Rihand

(Rihand–Delhi HVDC Scheme)
1. kVA Rating : 315,000 kVA, Single phase

three winding
2. Rated voltage
400 kV
Line Winding (HV) :
3
213 kV
Valve Windings (LV) : star and 213 kV delta
3
3. HV-LV impedance at
principal tap : 18%
4. Connections : Star/Star-Delta
5. Tappings : 24 steps, On load –12.5% to +17.5%
6. Type of cooling : OFAF
7. Terminals : HV : AC Bushing
LV : DC Bushing
8. Insulation levels
8.1 Impulse HV : 1300 kVp Lightning
1080 kVp Switching
LV Star : 1550 kVp Lightning
1290 kVp Switching
LV Delta 1175 kVp Lightning
980 kVp Switching
8.2 Induced voltage test
8.2 with PD : 440 kV
8.3 Ling duration DC voltage
8.3 test associated with
8.3 partial discharge : 800 kV for one hour
8.3 Direct voltage polarity
8.3 reversal test associated
8.3 with partial discharge : 568 kV
good tensioning arrangement and also adoption of stabili-

zation process of the windings.
— Drying process of the transformer, better quality processed
oil, soaking time also govern the quality of convertor trans-
former in terms of other electrical parameters like tan-delta
value, insulation resistance partial discharges, etc.
— Finally, cleanliness of the working environment, components

and accessories used also play their role in the manufacturing
process of convertor transformers.
22.10 Tests
In order to simulate the actual operating conditions, the convertor

transformers are subjected to both normal AC and DC dielectric
tests. Recommendations for various test levels and sequence of tests
for valve windings are given in Electra no. 46, Electra no. 157 and
IEC 61378-2. The AC windings are tested in the usual manner in
accordance with IEC 76/IS: 2026. All other routine and type tests are
the same as for an AC transformer.
The following tests are performed on the valve winding to assess
its ability to withstand various stresses due to DC, AC impulse,
polarity reversal, or short time DC over-voltages.
22.10.1 Long Duration DC Voltage Test
This test is performed to demonstrate the ability of the transformer

to withstand stresses due to continuous DC voltage. A DC voltage
of positive polarity is applied for 60 minutes. The windings not
under test are short-circuited and grounded together with the
tank. The level of the test voltage (UDC) for one hour is obtained as
follows:
UDC = 1.5 [(Z – 0.5)Ud + 0.7 Uvo]
where, Z : The 6 pulse bridge number starting at ground level.
Ud : Nominal direct voltage across one 6 pulse bridge.
Uvo : No load voltage (rms) between the valve winding phase
to phase terminals.
During the last 10 minutes of the test, measurement of DC partial
discharge is made. The number of discharge pulses exceeding 2000
pico-coulombs should not exceed 10. If the number of pulses exceed
10, the test is prolonged by another 10 minutes, and if after three
ten-minute observation periods the number of pulses exceed 10 per
period, the transformer is rejected.
22.10.2 Polarity Reversal Test coupled with

DC Partial Discharge Test
To demonstrate the ability of the transformer to withstand stresses

due to polarity reversal or other rapid DC voltage changes polarity
reversal test coupled with DC partial discharge monitoring is
performed.
A DC voltage of negative polarity is applied for at least two hours,
at the end of which the polarity is reversed within two minutes, and
the reversal voltage is applied for a further 30 minutes. During the
last 29 minutes, the number of pulses exceeding 2000 pico-coulombs
should not exceed 30. If the number of pulses exceed 30, the voltage
polarity is changed and the whole test is repeated. If the number
of pulses measured in the final 29 minutes again exceeds 30, the
transformer is deemed to have failed the test. The level of the test
voltage (Upr) is obtained as follows:
Upr = 1.25 [(Z – 0.5) Ud + 0.35 Uvo]
where Z, Ud, and Uvo are as mentioned earlier under the long
duration DC voltage test. During the test, the windings not being
tested are short-circuited and grounded together with the trans-
former tank.
22.10.3 AC Applied Test Along with PD Measurement
This is a good test to develop higher stresses in oil ducts. The

recommended test level is 1.5 times the crest operating level for the
convertor transformer. The test level would be as follows:
1.5 2
Vac applied = [(Z – 0.5) Ud + ¥ Uvo]
2 3
where Z, Ud, and Uvo are as mentioned earlier under the long
duration DC voltage test
The AC voltage is applied for one hour along with partial discharge
measurement.
22.10.4 Switching Impulse Test
Switching impulse voltage is applied to the valve winding terminals

connected together. The other winding terminals are short-circuited
and grounded together with the tank. The polarity is negative and
the wave shape is kept in accordance with the prevailing standards.
22.10.5 Lightning Impulse Test
Lightning Impulse voltage is applied to each valve winding terminal

in turn. The other valve winding terminal and all other winding
terminals are grounded together with the tank.
REFERENCES
1. Wahlstrom, Bo., (WG 12.02), Voltage Tests on Transformers and
Smoothing Reactors for HVDC Transmission, Electra No. 46, May
1976.
2. Watanabe, M., T. Obata, E. Takahashi, Development of Convertor
Transformers and Smoothing Reactors, Hitachi Review, Vol. 34,
1985, No. 5.
3. Mannhein, H., Bold, Static Convertor Transformer for the HVDC
Transmission Link Cabora Bassa, Apollo, Brown Boveri Rev. 7, 76.
4. Beletsky, Z.M., I.D. Voevodin, A.K. Lokhanin, Transformer
Equipment for DC Transmission Lines, World Electrotechnical
Congress, June 1977.
5. A. Lindroth et al. “The relationship between Test and Service
Stresses as a function of Resistivity Ratio for HVDC Converter
Transformers and Smoothing /Reactors” CIGRE Working Group
report WG 12/14.10, Electra No. 157, December, 1994.
6. Part 2 : Transformers for HVDC applications, IEC 61378-2, first
edition, 2001–02.
C HAPTER 23
Controlled Shunt Reactor

S.C. Bhageria
J.S. Kuntia
Long distance EHV and UHV lines experience fundamental frequency

temporary over-voltages under no-load and light-load conditions,
and also due to load rejections as per the Ferranti-effect. Such over-
voltages can stress the equipment connected to the lines, and need
to be controlled. It is a common practice to provide fixed shunt reactors
in these lines to compensate the capacitive line charging current,
and thereby control the over-voltages under no-load, light-load,
and load rejection. These shunt reactors also control the switching
over-voltages to some extent.
However, fixed shunt reactors have some major disadvantages.
In loaded power lines, these reactors continuously absorb rated
reactive power, increase the surge impedance of the line, and thereby
reduce the surge impedance loading (SIL) level, that is, the level at
which a flat voltage profile along the line can be achieved. These
permanently connected fixed shunt reactors also consume active
power, which is a continuous loss to the system. The disadvantages
associated with permanently connected fixed shunt reactors call for
the development of a controlled shunt reactor, that is, a reactor in
which reactive power absorption can be varied under changing load
conditions.
23.1 Controlled Shunt Reactor (CSR) Principle
The basic principle of the CSR is to control the reactive power

by using a thyristor valve, which can provide the necessary speed
of switching and control by means of firing angle control. As it is
not economical and practical to use the thyristor valves at high
transmission voltages like 400 kV, the thyristor valves are connected
to the controlled shunt reactor transformer (CSRT) secondary side.
A CSRT with primary winding connected to high voltage transmis-
sion lines, and secondary winding (control winding) connected to
thyristor valves, forms the basis of CSR. Basically, a CSR consists of
a high impedance transformer (CSRT) controlled by an antiparallel
pair of thyristor valves. The impedance of a CSRT can be controlled
by varying the firing angle of the thyristor pair through a controller.
The controller is a logical and programmable device, which generates
firing pulses based on the input signals. Figure 23.1 explains the
underlying principle of the CSR. When the thyristor valve is not con-
ducting the control winding is almost open. In this case, only a very
small current (very low MVA) flows through the primary winding
because the reactive impedance of a CSR is very high. On the other
hand, when the thyristor is fully conductive (i.e. control winding is
short-circuited) the inductive impedance of the CSR is nearly 100%.
In this case, the current through the primary winding is maximum
(nominal MVA) and is approximately equal to the short-circuit
current when the secondary of the CSRT is short-circuited.
Transmission line
CSR CT Voltage transformer

Line CT
IL
Main winding (100%)
CSR controller
Secondary winding
IR
Status
inputs
Thyristor
valve
Firing Firing
pulses circuits
Compensating winding
in delta
Figure 23.1 Principle of controlled shunt reactor.
The CSR generally consists of a controlled shunt reactor trans-

former with approximately 100% impedance, thyristor valves, a
Controlled Shunt Reactor 541
controller, a neutral grounding reactor, the necessary circuit break-

ers, and other auxiliaries. Figure 23.2 shows the general arrange-
ment of the controlled shunt reactor (CSR) scheme.
BHEL has successfully developed and commissioned a 50 MVAr,
400 kV class controlled shunt reactor (CSR) at the 400 kV Itarsi
Substation of Power Grid Corporation of India Ltd.
This type of reactor has been developed for the first time in
the country. The CSR is not merely a variable shunt reactor—its
advantages are much more than the shunt reactor. The development
has been possible due to excellent support provided by the customer
M/s. PGCIL. Figure 23.3 shows the installed CSR and Fig. 23.4 shows
a panaromic view of the CSR installed at the Itarsi site.
23.2 Controlled Shunt Reactor Transformer (CSRT)
The CSRT is a power transformer with nearly 100% impedance

between primary and control winding (secondary winding). In other
words, when the control winding is short-circuited with rated voltage
applied on the primary winding, the reactive power flow will be
similar to the power flow in a shunt reactor of similar rating
connected to the same rated voltage.
One end each of the star connected primary winding of CSRT is
directly connected to three phases of line. The neutral end of the
primary is grounded through a reactor. The star connected control
windings with grounded neutral are connected to the thyristor
valves, and can be controlled independently. It also consists of
the compensating winding (tertiary) connected in delta for the
compensation of the 3rd and its multiple harmonics.
The CSRT, when its secondary is kept open, draws only the mag-
netizing current, and consumes negligible reactive power. Another
important aspect of the reactor transformer is its magnetisation
characteristic, which remains practically linear as the main flux path
under loaded condition is mostly through the air. Hence, even under
sustained over-voltage conditions, the problem of core saturation
does not occur.
The specifications and design of the CSRT depend on the perfor-
mance specifications of the CSR as a system. The construction of the
CSRT is similar to that of a power transformer, which consists of a
closed magnetic core, and co-axial windings.
400 kV transmission line
R
542
Y
CB B
Line CTS for controller
R Y B
400 kV CVT
isolator &
earth switch VR
VY For controller
VB
Thyristor valves
N
Reactor
CTS for transformer
controller R Y B
Tertiary Second-
Primary
Transformers, 2/e
ary
Neutral
grounding
reactor
Vacuum
circuit breaker
By pass
chokes
Figure 23.2 General arrangement of controlled shunt reactor (CSR) scheme.

Controlled Shunt Reactor
Figure 23.3 CSR installed at PGCIL itarsi site.

543
544
Transformers, 2/e
Figure 23.4 Panoramic view of CSR installation at itarsi.

23.3 Special Features of the Controlled Shunt

Reactor Transformer
Some of the other important aspects of the design and construction

of the CSRT are mentioned below:
(a) Voltage and current rating of winding: The voltage and current
rating of the primary winding is decided on the basis of the system
voltage and the reactive power requirement. The selection of sec-
ondary voltage and current depends on the availability, suitability,
and overall economy of thyristor valves. Usually, the compensating
winding voltage rating will be the same as that of control winding,
and its capacity will be adequate for suppression of harmonics.
However, for special application of the CSR viz. compensation of
reactive power, the required capacity of compensating winding may
be as high as 100% of the CSRT rating.
(b) Winding disposition: The requirement of 100% impedance
between primary and control windings can be achieved by a suitable
selection of the voltage per turn and the radial gap between these
windings. The control winding being the low voltage winding, is
generally placed next to the core. The compensating winding is
generally placed between the primary and control windings. This not
only offers an economical solution, but also has several technical
advantages. In this position, the compensating winding is closely
coupled with the control winding for very good compensation of
harmonics generated in the control winding. On the other hand, the
compensating winding is loosely coupled with the primary winding,
and helps to reduce the effect of circulating currents on the primary
winding. This is particularly important for making the CSR system
compatible for single phase auto-reclosure. The interwinding gaps
play a very important role in achieving the desired performance
and optimum design of CSRT. Figure 23.5 shows the disposition of
winding in the CSRT.
(c) Stray loss control: As mentioned earlier the CSRT is a power
transformer with approximately 100% impedance between the pri-
mary and control windings. The very high leakage impedance, and
special disposition of the windings make stray losses increase many
fold in the case of the CSRT. When the CSR is required to provide
full reactive power, the control winding is completely shorted. In
Top yoke
C
C o P
C o m r
o n p i
r t e m
e r n a
o s r
l l a y
i t
m i
b W n W
D g D
G W G
D
G
Bottom yoke
Figure 23.5 Winding disposition in controlled shunt reactor transformer.
this condition, there is no flux coupling between the primary and

control windings. On the other hand, as the primary winding is
connected to the nominal voltage, all the flux linking the primary
winding is disposed in the radial gap between the primary and con-
trol windings. In other words, during the nominal regime of the CSR,
there is no flux in the main limb of the CSRT, but is disposed in the
interwinding gap of the primary and control windings. This flux is
also called the leakage flux, and because of the approximately 100%
impedance between the primary and control windings, the leakage
flux emanating through the annular space between the windings re-
sults in very high stray losses, that is, of the order of 630% of the I2R
loss, without yoke shunts, wall shunts, and non-magnetic clamping
structure. Figure 23.6 shows the magnetic field distribution of the
newly developed CSRT without yoke shunts and wall shunts. From
the field plot, it is evident that in the absence of yoke shunts, there is
maximum fringing of flux at the top and bottom. This stray flux is
responsible for creating very high losses in the adjacent structure
and tank. It may be noted that such high stray losses in the adjacent
Figure 23.6 Magnetic field plot without yoke shunts and wall shunts.
structures and tank may not allow satisfactory operation of the

CSRT, and may lead to temperature rises beyond guaranteed values.
The generation of excessive dissolved gasses etc. will alter the oper-
ating characteristics of the CSRT, and are not acceptable. Special
methods like using yoke shunts and wall shunts, and providing them
on the strategic locations with the help of accurate magnetic field
analysis were employed to control the stray losses. It is important to
note that successful implementation of these methods could reduce
the stray losses to about 50% of I2R losses, compared to stray losses
of 630% of I2R losses without these methods. Figure 23.7 shows the
magnetic field plot of the CSR provided with these methods. From
the field plot, it is evident that the flux path has become almost par-
allel to the windings, and fringing at the ends has reduced to a great
Figure 23.7 Magnetic field plot with yoke shunts and wall shunts.
extent. A very high percentage of flux has been collected by yoke

shunts and pumped to the core. Similarly, the stray flux near the
tank is passing through the wall shunts, and the entry of flux in the
tank wall has been reduced to a great extent.
It may be noted that further reduction of stray losses to the extent
of approximately 10–15% of I2R loss is possible by additional use of
non-magnetic structures for the end frame and clamp plate.
It is important to mention here that the actual measured losses
were approximately in line with the expected losses with the provi-
sion of newly developed yoke shunts and wall shunts.
23.4 Controlled Shunt Reactor Much More

than a Shunt Reactor
The CSR is not merely a substitute of the shunt reactor—its advan-

tages are much more than that. The major advantages are:
(a) Fully controllable reactive power: The CSR provides
reactive power support to the line only when it is required,
thus reducing the continuous reactive power drawn as in the
case of fixed shunt reactors.
(b) Reduction in dynamic over-voltage limits: The CSR
automatically goes out of circuit during increased line loading
conditions, thus eliminating the need to limit the reactive
compensation limit to 60% which is the present practice in
Indian grids. This feature can be used to provide higher
compensation on the lines to limit the power frequency
dynamic over-voltages. This shall help in reducing the
over-voltage requirements of all the substation equipment,
resulting in overall economy.
(c) Increased power carrying capacity of lines: The CSR
control system automatically takes it out of circuit under full-
load condition of the lines. Thus, an increase in the power
transmission capacity is achieved. In comparison to the lines
compensated with the fixed shunt reactor, the transmission
lines provided with CSR can enhance the power transmission
capacity by 25–30%.
(d) Fast response: The CSR can respond with a short response
time of 10 milliseconds for any sudden over-voltage etc., and
can come into the circuit with full reactive power support.
This feature eliminates the risk involved in taking out the
fixed reactors from the circuit.
(e) Full compatibility to single phase auto-reclosure: The
fast response time of the CSR makes it suitable to respond to
fault conditions like single line to ground and three phase
faults. The control system senses the fault, and brings the
reactor to full conduction within a short time, thus providing
full compatibility for single phase auto-reclosure. The need
for NGR is similar to that of a fixed shunt reactor.
(f) Ability to connect directly to EHV level: The controlled

shunt reactor transformer can be directly connected to the
EHV level, as against an SVC system, which requires an
intermediate transformer in addition to the controllable
reactor.
(g) Minimum harmonics: The continuously controlled CSR is
provided with small filters across the delta connected
compensating winding. This is to limit harmonics generated
due to variation of the firing angle of the thyristor valves in
accordance with the reactive power requirement. The total
harmonics generated due to thyristor firing will be less than
2–3% with respect to the nominal capacity of the CSR.
23.5 Conclusion
Thus, the CSR offers various advantages like fully controllable

reactive power, reduction in dynamic over-voltages, increased power
carrying capacity of lines, fast response, full compatibility to single
phase auto-reclosure, economy of size, minimum harmonics etc.
It is expected that the CSR will provide an economical solution for
stable and strong transmission networks. Controlled shunt reactors
placed at suitable points in a line can make the line carry surge
impedance loading level of power with nearly 1.0 P.U. voltage
profile, and theoretically make it possible to make AC power get
transmitted over infinite long distances.
REFERENCES
1. S.C. Bhageria, G.N. Alexandrov, S.V.N. Jithin Sunder, and M.M.
Bhaway. "Reactor Transformer for Controlled Shunt Reactor
Applications". International Conference, Power Transformers 2000
by CBIP.
2. S.V.N. Jithin Sundar, S.C. Bhageria, C.D. Khoday, Amitabh Singhal,
A.K. Tripathi, G.N. Alexandrov, M.M. Goswami, I.S. Jha, Subir Sen,
V.K. Parashar "Controlled Shunt Reactor—A member of FACTS
family". Eleventh National Power System Conference (NPSC-2000).
C HAPTER 24
Designing and Manufacturing

—A Short-circuit Proof
Transformer
T.K. Ganguli
S.K. Gupta
The short-circuit withstand capability of a power transformer is

defined as the ability to withstand the full asymmetrical short-circuit
currents without impairing its suitability for normal service
conditions. Steady increase in unit ratings of transformers and
simultaneous growth of short-circuit levels of networks have made
the short-circuit withstand capability of the transformer one of the
most important aspect of its design.
The short-circuit (SC) test is a very stringent test that involves
symmetrical short-circuit currents of the order of 6–7 times the
rated current and peak asymmetrical current as high as 15–18 times
the rated current. The basic formula for force acting on a current
carrying conductor in a magnetic field at any instant is given by:
F = BIL Newtons,
where,
B = Flux density in Tesla
I = Current in Amps
L = Length of conductor in metres
This means that resulting forces shall be tremendously high since
they shall increase in square of the current. These forces shall be
exerted directly on all current carrying parts of the transformer in a
magnetic field, i.e. windings, leads, terminal gear, etc. Short-circuit
forces are of pulsating nature since short-circuit current and flux
are sinusoidal. Therefore, time variable stresses are applied to the
affected structures. These forces are very high during the first
asymmetrical peak and come down to the symmetrical rms value.
24.1 Forces During Short-circuit
The forces generated during short-circuit are transferred to all parts

mechanically coupled to winding assembly. The overall effect of
these forces has been elaborated below:
24.1.1 Radial and Axial Forces on Winding
Due to leakage field distribution during SC, heavy radial and axial
forces shall be generated inside windings. This has already been
elaborated in Chapter 7. The inner windings under compressive
stress have a tendency to fail under buckling as described below.
Buckling Phenomena
This has already been explained in chapter 7. The buckling strength
of a transformer is dependent upon the elasticity of the material,
that is, work hardening, conductor radial thickness, base cylinder,
etc. Buckling is a very common mode of failure on a transformer
subjected to SC test.
24.1.2 Forces on Winding Leads
Winding leads are normally perpendicular to winding conductors in

most the cases, as in Fig. 24.1 below.
The leakage flux and current in lead sections shall result in a
force perpendicular to the lead length. For inner windings which are
under compressive stress, the tendency of winding leads shall be to
tighten the screw of the coil by twisting itself towards a smaller
diameter winding. This is called the spiralling effect and is one of the
common causes of failure in helical and spiral windings. Spiralling
has been further explained below.
Spiralling
Spiralling involves a marked deviation of the spacer rows from their
original, vertically aligned arrangement. The spacers undergo a
progressive displacement, starting from the lower end to the upper
end of the winding generally combined with a pronounced relative
shift of the last bottom and top turns with respect to turns nearby,
Designing and Manufacturing 553
Forces on lead
Top lead
section
sections causing
screwing effect
Winding
conductors
Bottom lead
section
Forces on lead
sections causing
screwing effect
Figure 24.1 Forces on lead sections during short-circuit spiralling effect.
the shift being larger for the top turn. This also results in bending of
the exit leads and crossovers.
24.1.3 Forces on Base Cylinder and Core
It has already been explained in Chapter 7 that the inside windings,

specially the LV winding and inner tap windings, are normally
subjected to inside radial stresses. These windings are normally
wound over the base cylinder. Although winding is considered as a
solid mass, radial stresses are still exerted onto the core through the
base cylinder and the insulation arrangement between the core and
winding. So, the base cylinder must be adequately designed to with-
stand short-circuit forces. The radial forces are almost uniform along
approximately two-thirds of the winding height. However, it starts
decreasing at winding ends due to radial bending of flux lines at ends
(see Fig. 7.1)
24.1.4 Forces on Winding Clamping Structure
It is evident that short-circuit results in very high axial forces in the

winding, to the order of tens or hundreds of tonnes for large rating
power transformers. Normally, the net cumulative axial force
on one winding is upwards, while on the other winding it is in the
opposite direction, that is, downwards.
A typical axial clamping structure of a large transformer is shown
in Fig. 24.2. The axial forces from winding during short-circuit are
transmitted to winding end blocks/block washer and top ring, and
then to the top yoke or clamping structure. The axial clamping
structure, pressing rings, etc. form the backbone of the transformer
for withstanding SC forces.
Clamping screws
End frame Top yoke
Top ring
Top end blocks/

block washers
Forces
towards
top HV
coil
Core
LV
coil Forces
towards
bottom
Bottom end blocks/
block washers
Bottom ring
Bottom yoke
Figure 24.2 Typical axial clamping structure and typical axial

forces pattern during SC test.
24.1.5 Terminal Gear Items and their Supports
The lead connections from the winding ends are taken out and
connected to form a star/delta inside transformer for a 3-phase
transformer and then connected with bushings at the top. This

involves a complete terminal gear which is supported by various
cleats and supports. During short-circuit leakage flux are produced
in the areas adjacent to the windings and approaching towards the
tank (refer to leakage field distribution in Chapter 7). Thus, forces
are generated in the leads:
(a) Due to the presence of leakage flux of winding. These forces
shall be in a direction perpendicular to the length of the lead.
(b) Due to the interaction of two current carrying conductors
either in the same direction (for example different parallels of
same winding), or in the opposite direction (for example LV
and HV leads). The resulting forces in the former case shall
be mutually attractive, while in the latter case, it shall be
mutually repulsive, as illustrated in Fig. 24.3.
I1 I2 I1 I2
Force of attraction Force of repulsion (current

(current in same direction) in opposite direction)
Figure 24.3 Forces in two current carrying conductors.
The above forces shall be exerted through these conductors on to

their support structure. Low voltage leads of large transformers
have very high short-circuit currents and the forces can be quite
significant. Tap leads also experience considerable force due to the
vicinity of a number of cables.
24.2 Various Considerations to Design a Transformer

Suitable for Short-circuit Duty
The design stage is a very crucial aspect for a short-circuit proof

transformer, and if proper care is taken during the design stage,
then the short-circuit forces can be controlled. Also, the withstand
capability of the various parts can be enhanced to a considerable
extent. Areas where proper attention is required are explained
below.
24.2.1 Radial Stress Withstand of Winding
Theoretically, radial stress in a winding can be reduced by reducing

the current density, or by increasing the impedance. However, there
are several practical limitations for a change in the above param-
eters. It is not always wise to reduce the current density as it is quite
uneconomical. Similarly, impedance is also guaranteed as per sys-
tem requirements and has to be within tolerances.
The radial withstand capability of a winding can be enhanced by
increasing the work hardening strength of the winding conductor, by
increasing radial thickness of the winding conductor, by increasing
the number and dimensions of supporting blocks, etc. However,
there are some practical constraints. For instance, very high work
hardening means a practical difficulty in manufacture of windings
and poor control over manufacturing of conductors. Similarly,
increasing the number and size of blocks would reduce the distance
between blocks, which may result in inadequate cooling surface
or insufficient distance for transposition/crossovers. Therefore, a
judicious selection of design of the winding is necessary for a
designer. Increasing the radial stress withstand would also improve
the buckling strength of the winding.
24.2.2 Axial Forces Withstand of Windings
Axial forces arise out of ampere turn imbalance between primary

and secondary windings, and due to the bending of flux lines at wind-
ing ends. The ampere turn balance along the height of the winding is
therefore one of the most important aspects for reducing the axial
force itself, which is the crux of the problem. Ampere turn imbalance
occurs due to unequal turns per disc, specially in disc windings or
due to unequal duct size distribution along the height of the winding.
A good designer always aims at reducing these inbalances by
minimising and adjusting the turns distribution and duct distribu-
tion so as to result in minimum possible axial forces.
Another aspect is to enhance the axial withstand capability of the
winding itself. This is possible by decreasing the width/thickness ra-
tio of the winding conductor, increasing the number and size of wind-
ing blocks, etc.
24.2.3 Top and Bottom Yoke Clearances
Due to design requirements, sometimes the top and bottom yoke

clearances are kept unequal. This is specially required if the HV line
lead is to be taken out from top. In this case, the only solution is to
enhance the axial force withstand capability of the winding.
24.2.4 Location of Winding
Whenever a winding is kept in the main flux due to the design and
client's requirement, it shall experience a heavy radial force due to a
very high component of axial force in this region (See Fig. 7.1 of
Chapter 7).
In this arrangement, the radial withstand capability of the wind-
ing must be prudently enhanced. The winding ends or sections of
winding near winding ends are the most critical area and require an
adequate supporting or locking structure to arrest their radial and
peripheral movement.
24.2.5 Anchoring and Locking of Leads
Winding ends/leads often become weak points, hence proper anchor-

ing of leads with adjacent sections/turns is essential. Anchoring is
very important for tapping winding since a number of leads are
taken out from the winding and it often becomes a weak point. A
proper locking arrangement to arrest the movement of tapping leads
is therefore very essential.
24.2.6 Design of Top Ring
The thickness of the ring is calculated based on the axial short-

circuit force, the number and size of clamping screws, the span
between two clamping screws, and the number and size of slots for
taking out LV and tapping leads. The slot in the top ring should be as
small as possible. The material of the top ring, its thickness and size,
its method of joining and the material strength should be ensured
properly before use. Also, additional pressure pads can be provided if
the span between two screws is too high.
24.2.7 Selection and Design of Clamping Screws

and Supporting Structure
The capacity and number of clamping screws/pressure pads must be

adequate to withstand the maximum axial force expected during
short-circuit. The clamping screw dimension, viz. its diameter,
length, and basic material is important. The clamping screw is tight-
ened and kept in vertical position to exert uniform pressure to the
windings.
Similarly, the complete support structure including the endframe
and clamp plate/pin, etc. must be designed to meet the axial force
arising during short-circuit.
24.2.8 Terminal Gear and Its Support
All terminal gear leads, especially of LV and tapping coil are to be

anchored and supported firmly. The supports must be suitable for
the forces arising during the short-circuit test calculated in line with
section 24.1. If required, the number of supports per unit length may
be increased. The thickness and size of the supports should also be
selected based on the short-circuit forces expected.
Spacing between leads is a very important aspect, since force in-
creases in inverse ratio of square of centre to centre distance be-
tween cables. Therefore distance pieces between two cables may be
inserted to maintain distance.
24.3 Manufacturing Aspects
Merely a good design cannot make a transformer short-circuit proof.

The manufacturing capability is also equally important in all stages
of manufacturing, as explained below.
24.3.1 Winding Manufacturing
Proper tightness of winding conductors, proper circular shape

of winding without any eccentricity, etc. is to be ensured during
manufacturing of winding. Otherwise, even with a good design
the transformer might fail under short-circuit. For example the
Buckling phenomena as explained in section 24.1 is very complex,
and is deeply influenced by the manufacturing process of winding.
Poorly wound windings exhibit reduced buckling strength whereas
stiff, well tightened and supported windings provide a significant
enhancement in the withstand capability against buckling.
24.3.2 Dimensional Stability
Meeting the dimensional stability of various parts, especially

winding is of utmost importance and must be well taken care of. The
winding height must be strictly in line with specified dimensions and
allowed tolerances.
24.3.3 Proper Curing of Epoxy Bonded Conductors
Proper curing of epoxy bonded conductors is one of the most impor-

tant aspects since the conductor cannot attain the specified strength
if proper curing is not done. The supplier's instructions must be
strictly adhered to.
24.3.4 Use of Calibrated Blocks for Windings
Preshrunk and calibrated clacks/blocks should be used to ensure

required winding height and for better dimensional stability of the
windings. Also, the quality of all base cylinders and end rings must
be ensured.
24.3.5 Stabilisation of Windings and Coil Assembly
All windings individually and the composite coil assembly are to be

properly stabilised as per the manufacturer's standard practice, to
ensure proper pressure on the winding assembly.
24.3.6 Crossover/Transposition Locations
In many cases, damage has been observed on the winding due to

shorting of conductors at transposition/crossover locations. These
areas are weak points, hence they should be critically taken care of
to avoid any damage due to sharpness etc. and additional insulation
should be provided in these areas for strengthening.
24.3.7 Tight Lowering of Coils
Tight lowering of coils is very essential in order to achieve a tight

assembly of windings. This shall ensure even distribution of force
and enhance the strength of the transformer under short-circuit. In
case ovality is observed, it should be packed with spacers/cylinders
for proper support.
24.4 Quality Aspects
Quality control during each stage of manufacturing is a must in

order to manufacture a transformer suitable for short-circuit. Each
stage, starting from raw material to various stages like winding,
core, assembly, terminal gear, and insulation items should be
subjected to strict quality control.
24.5 Conclusion
Unless design, manufacturing, and quality aspects are fully taken

care of, it is not possible to manufacture a transformer suitable for
short-circuit. The behaviour of a large power transformer during
short-circuit is quite complex and involves a number of other factors
like inner friction in the windings, stability in structural materials,
springing and damping within the winding, effects of oil motion, etc.
The selection of various fittings, i.e. tapchanger, bushing etc. should
be done so that it is able to withstand the stresses caused during
short-circuit.
Thus for designing and manufacturing a short-circuit proof
transformer all aspects, viz. design, material selection, selection of
its fittings and the manufacturing processes are to be carefully
considered and all minor aspects including support structures, etc.
are to be verified in detail for adequate strength.
C HAPTER 25
High Voltage Condenser

Bushings
R.K. Agrawal
Aseem Dhamija
25.1 Introduction
Bushing is one of the most important components that are fitted to

the electrical equipments like the transformer, the switchgear etc. It
is an insulating structure for carrying the HV conductor through an
earthed barrier.
The bushing has to
(a) Carry the full load current.
(b) Provide electrical insulation to the conductor for working
voltage and for various over-voltages that occur during
service.
(c) Provide support against various mechanical forces.
25.2 Classification of Bushings
Bushings are classified according to the following factors:
25.2.1 Application or Utility

(a) Alternator Bushing
AC generators require bushings up to 33 kV, but 22 kV is more
usual. With modern alternators, current ratings up to 20,000 A are
required.
High Voltage Condenser Bushings 563
(b) Bushings for Switchgear

In the switchgear, bushings are to carry the conductors through the
tank wall, and support the switch contacts.
(c) Transformer Bushings

Transformers require terminal bushings for both primary and
secondary windings. In some cases, a high voltage cable is directly
connected to the transformer via an oil filled cable box. A bushing
then provides the connection between the cable box and transformer
winding.
(d) Wall or Roof Bushing

In recent years, many sub-stations for 132 kV and above, in unfavor-
able situations have been put inside a building. For such applications
wall/roof bushings are used.
(e) Loco Bushings

These bushings are used in freight loco and AC EMU transformers
for the traction application.
25.2.2 Non-condenser and Condenser Bushings

(a) Non-condenser Bushing
In its simplest form, a bushing would be a cylinder of insulating
material, porcelain, glass resin, etc. with the radial clearance and
axial clearance to suit the electric strengths. The voltage is not
distributed evenly through the material, or along its length. As the
rated voltage increases, the dimensions required become so large
that this form of bushing is not a practical proposition. The concen-
tration of stress in the insulation and on its surface may give rise
to partial discharge. This type of bushing is commonly used as
low-voltage bushings for large generator transformers.
(b) Condenser Bushing

In this type, the conducting cylinders are inserted in the insulation
to divide the wall thickness into a number of capacitors. In this way,
the voltage distribution in the material and along its surface can be
controlled.
Ball or earthed barrier
Live
conductor
Paper
insulation
Non-condenser bushing
equipotential cylinders
Condenser bushing
(a) Ungraded 100 % (b) Graded
75 %
50 %
75 %
50 %
25 %
25 %
0% 0%
Voltage distribution with and without capacitance grading
Figure 25.1 Condenser and non-condenser bushing.

25.2.3 Insulating Material
The insulating material of bushings is usually paper-based with the

following most common types:
(a) Synthetic Resin Bonded Paper (SRBP)

In SRBP bushings, one side of the paper is coated with resin which is
cylindrically wound under heat and pressure inserting conducting
layers at appropriate intervals. However, use of SRBP bushings is
limited to low voltages. There is also the danger of thermal instabil-
ity of insulation produced by the dielectric loss of the resins. The
SRBP insulation is essentially a laminate of resin and paper which is
prone to cracking. Moreover, paper itself will include air which will
cause partial discharges even at low levels of electrical stress.
(b) Oil Impregnated Paper (OIP)

OIP insulation is widely used in bushing and instrument transform-
ers upto the highest service voltages. In the manufacturing process,
the kraft paper tape or sheet is wound onto the conductor.
Aluminium layers are inserted in predetermined positions to build
up a stress-controlling condenser insulator. The condenser layer
may be closer together, allowing higher radial stress to be used. The
bushing is fully assembled before being vacuum impregnated in
order to contain the oil.
(c) Resin Impregnated Paper (RIP)

RIP bushings are wound in a similar manner as OIP. The raw paper
insulation is then kept in a casting tool inside an auto-clave. A
strictly controlled process of heat and vacuum is used to dry the
paper prior to impregnation with epoxy resin. The resin is cured
under a heat and pressure cycle. The resulting insulation is dry, gas
tight, and void free.
25.3 Design of Bushing
In its simplest form, a bushing consists of a central conductor embed-

ded in a cylindrical insulation material having a radial thickness
enough to withstand the voltage. The design of a bushing depends on
the various voltages and over-voltages that it has to come across
during service. In order to study the factors which influence the

design of a bushing, it is convenient to consider a transformer
bushing with one end in the air, and the other in oil. The important
factors which effect the design are:
(a) Air-end Clearance

The air-end clearance has to be sufficient to meet the specified
over-voltage tests. It is also determined by the creepage distance,
and the proportion of it that is protected from the rain. Having
determined the air-end length, the air-end dimension of the internal
condenser can be determined. It is not necessary to grade 100%.
Internal grading of 70% or less will give adequate surface grading for
large bushings.
(b) Oil-end Clearance

As internal breakdown unlike air flashover, is more severe, specifi-
cations, therefore, demand an internal breakdown with a sufficient
margin (about 15%) above the air withstand value. Both power
frequency, and impulse voltage withstand tests have been used to
specify this characteristic.
(c) Number of Condenser Layers

The number of partial condensers is so chosen that the test voltage
of each partial condenser should be between 10 kV to 15 kV. If more
foils are introduced, it will cause too many folds and weaken the
bushing. Also, will be air introduced in the folds, complicating the
manufacture of bushing of high voltage class.
(d) Length of Earth Layer

The length of the earth layer of a bushing is usually determined
by the accommodation required for current-transformers, or by
mounting considerations, though in some cases it may be allowed to
assume its optimum dimension in relation to the radial dimensions.
The ratio of Length of first foil (L1) and Length of nth foil (Ln) may be
taken between 3 to 4. This ratio is denoted by a
(e) Radial Gradients and Diameters

The radial gradient is limited by the necessity for avoiding damage
by discharges at the power-frequency test voltages, whether one
minute or instantaneous.
If the ratio of the earth layer diameter to that of the conductor

rn /ro, is denoted by b, the stresses at the HV end and the earth
voltage end will be equal, if the product of a and b is unity.
However, it is not always possible to achieve this value. Hence a
and b can vary from 0.8 to 1.2
If a. b = 1, then LnDn = L1. D0
(f) Equipotential Layer Position

After determining the dimensions of the inner and outer layers of
the condenser, the position of the other layers can be calculated. The
basis of the design of the condenser bushing is generally equal
partial capacitances, which mean equal voltage on them and equal
axial spacing between the end of layers.
25.4 Constructional Details and Main Parts of Bushing
25.4.1 Core
The core of bushing consists of a hollow or solid metallic tube, over

which high grade electrical kraft paper is wound. For condenser
cores, conducting layers of metallic foil are introduced at predeter-
mined diameters to make uniform distribution of electrical stress.
The winding of the condenser core is done in a dust-free chamber.
The core is then processed; this comprises of drying in a high degree
of vacuum (0.005 mm), and then impregnating with high quality,
filtered and de-gassed transformer oil.
25.4.2 Porcelain
Bushing for outdoor applications are fitted with hollow porcelain in-
sulators. The OIP bushings are provided with insulators, both at air
and oil ends, thus forming an insulating envelope, and the interven-
ing space may be filled with an insulating liquid or another insulat-
ing medium.
The function of an insulator is to resist flash over in adverse con-
ditions. This is determined by
(a) the profile of the dielectric.
Top terminal
Oil level Top cap

indicator
Top porcelain
Test tap
Flange
Bottom
procelain
Bottom stress
shield
Figure 22.2 Main parts of bushing.

(b) the mounting arrangement of the insulator, i.e, vertical,

horizontal, or inclined.
(c) the properties of the surface, i.e, hydrophobicity, toughness
etc.
For insulators used in the air-end side, a minimum specific creepage
distance is required for different pollution levels, as given in
Table 25.1.
Table 25.1
Pollution level Creepage distance

I Light (areas without industries, agricultural 16 mm/kV
or mountain areas)
II Medium (areas with industries not producing 20 mm/kV
pollution smoke, low density of houses, areas
not exposed to the sea coast)
III Heavy (areas with high density of industries, 25 mm/kV
high density of heating plants producing
pollution, areas close to the sea)
IV Very heavy (areas close to the coast and exposed 31 mm/kV
to sea spray, desert areas, areas exposed to strong
winds carrying salt and sand etc.)
Based on the above information, the shapes of porcelain sheds are:

(a) Plain
(b) Anti-fog
(c) Alternate
25.4.3 Top Cap
This is a metallic housing for the spring pack. It serves as an in-built

oil conservator to cater for oil expansion, and has an oil level indica-
tor. In many cases, it also serves the purpose of a corona shield.
25.4.4 Mounting Flange
This is used for mounting the bushing on an earth barrier, such as a

transformer tank or a wall. It may have the provisions for following:
Plain or normal shed
Antifog shed
Alternating shed
Figure 25.3 Shapes of porcelain sheds.

(a) CT accommodation length

(b) Rating plate giving the rating and identification details of
bushing
(c) Test tap
(d) Oil drain plug for sampling of oil
(e) Air release plug
The design of the flange and top cap is such as to minimize the loss
due to hystersis and eddy current effects. When heavy currents are
being carried, this loss raises the temperature of the flange and top
cap to a noticeable extent. For heavy currents, ordinary cast iron
material cannot be used, hence non-magnetic materials such as
stainless steel or aluminum are used.
25.4.5 Test Tap
The test tap is provided for measurement of the power factor and
capacitance of the bushing during testing and service. The test tap
is connected via a tapping lead to the last condenser foil of the
core within the bushing. During normal service, this tapping is elec-
trically connected to the mounting flange through a self-grounding
arrangement.
25.5 Testing of Bushing
To prove the design and quality of manufacture, bushings are sub-

jected to type tests and routine tests as given in IS 2099 and IEC 137.
Where the type tests are done to prove design features of bushing,
routine tests check the quality of individual bushing.
25.5.1 Tan Delta and Capacitance Measurement
This test is probably the most universally applied test for all types of
condenser bushings. The bushing is set up as in service connected to
one arm of the Schering bridge. The voltage is applied in increasing
steps, up to the rated voltage. Capacitance and tan delta values are
recorded for each voltage. (For bushings, power factor and tan delta
values may be regarded as identical). Tan delta indicates the degree
of processing of the condenser core in OIP bushings and also the

moisture content.
25.5.2 Dry Power Frequency Voltage Withstand Test
This is the most common routine test used for various classes of
electrical equipment. A specified power frequency voltage is applied
for one minute. Under wet conditions, the power frequency test is
a type test, and is done only for bushings having rating 300 kV and
below.
25.5.3 Partial Discharge Test
Partial discharges are the localized electrical discharges within

the insulation system, restricted to only one part of the dielectric
material, thus only partially bridging the electrodes. They are caused
due to:
(a) voids and cavities present in the solid and liquid dielectrics.
(b) surface discharges that appear at the boundary of different
dielectrics.
(c) corona discharges, if strongly non-homogenous fields are
present.
The continuous impact of discharges in solid dielectrics form dis-
charge channels called TREEING. Every discharge event deterio-
rates the material by the energy of impact of high electrode ions,
causing many types of chemical transformation.
25.5.4 Impulse Voltage Withstand Test
Lightning impulse is a type test, and is applicable for all types of

bushings. The bushing is subjected to 15 full wave impulses of posi-
tive polarity, followed by 15 full wave impulses of negative polarity of
the standard waveform 1.2/50 ms.
For bushings of rated voltage equal to or greater than 300 kV,
the switching impulse voltage test is applicable. The impulse tests
simulate more closely than the power frequency withstand test, the
over-voltages likely to be seen by the bushing in the service.
25.5.5 Thermal Stability Test
The theory of thermal stability states that at an elevated operating

temperature, stable thermal equilibrium is assured only when a
maximum value of the sustained voltage characteristic of a particu-
lar bushing is not exceeded. The magnitude of this voltage serves as
a representative measure of the thermal stability. The magnitude
depends solely on the quality of the dielectric, its ambient tempera-
ture, and the manner in which it is internally cooled. The dimension
of the body has no role to play. It has been considered unnecessary
to specify thermal stability tests for OIP bushings owing to low
dielectric loss. However, it is necessary in large bushings (greater
than 300 kV) with high current, to pay attention in the design to the
dissipation of the conductor losses which may be several times the
dielectric loss.
25.6 Factors Affecting the Performance of Bushing
25.6.1 Dielectric Loss in Bushing
The principal factors governing the dielectric loss in OIP bushings

are the quality of oil, and the manner in which the bushing is
processed. The quality of dielectric is expressed not so much by the
absolute magnitude of the loss as by their constancy with respect to
time. The risk of ageing is considerable in OIP bushings. The oil
molecule itself does not possess very high resistance, and it is
particularly sensitive to high temperature. Low viscosity of oil
allows particles of foreign matter and ageing products to migrate
to the points at which the field concentration is greatest in an
irreversible manner.
25.6.2 Internal Corona
Corona discharges attack the dielectric, and thus weaken the bush-
ing electrically. The corona discharges crack the oil molecule. This
produces flammable gases and unsaturated molecular residues,
which accelerates the deterioration process. Even low discharge
intensities can result in a breakdown after only a short time. Hence,
it is essential that corona should not be allowed to persist for a long
time.
25.6.3 Atmospheric Pollution
Failures are also caused due to pollution. When the layer of pollution
on an insulator becomes wet, its resistance falls and a leakage current
flows over the surface of the insulator. The density of leakage
current is high in some parts, usually in the narrowest portions of
the insulator. These regions dry more quickly than the rest of the
surface, and their resistance increases. This results in the formation
of dry bands around the insulator. The voltage on the insulator is
virtually impressed across the dry bands, which results in a flash
over across the insulator in due course of time.
25.6.4 Temperature Rise of Joints
The current carrying path in a bushing generally has 3–4 joints from
the transformer winding to the top terminal. These joints are either
soldered, brazed, or bolted. Any looseness will cause a rise in the
temperature. If this is not timely checked, the excess heat will pass
to the insulation body, ultimately causing damage to the bushing.
25.7 Condition Monitoring at Site
On principle, all solid and liquid insulating materials are exposed to a

certain degree of ageing due to various stresses in service. Continu-
ous monitoring at the field provides necessary data of insulation
behaviour. Safe running of the equipment can be ensured if corrective
action is taken in advance, in case of any abnormality. The following
can be monitored on site on a regular basis to avoid any mishap:
25.7.1 Monitoring of Dissipation Factor and Capacitance
The tan delta or dissipation factor is considered to be the most

dependable parameter in bushing to assess its condition. The
increase in value of tan delta as well as capacitance gives very useful
information about any insulation deterioration. The measurement
at site can be done with a precision Schering bridge at 10 kV by
simultaneously balancing the bridge for capacitance and tan delta.
The test is conducted in Ungrounded Specimen Test (UST) mode,
which eliminates the losses going to the grounded portions of the
bushing.
The measured values of tan delta and capacitance can be

compared with the factory test results. Though the tan delta value
measured at site may not match with the factory results, however,
its limiting value for OIP bushings as per standards is 0.007. The
value of capacitance may vary from manufacturer to manufacturer,
depending upon geometry or construction of the equipment.
Every time the measurements are taken, the value should be
compared with the previous readings. If the bushing is more than 5
years of service, an increase in tan delta up to 1% can be allowed.
Increase in capacitance value by more than 5% of the original value
should always be referred to the manufacturer.
25.7.2 Monitoring of Hot Spots
A thermo-vision camera can be used to monitor the temperature of

the bushing, particularly near the top terminal. Abnormally high
temperature indicates the following:
(a) Improper engagement of top terminal with pull through
connector.
(b) Looseness of terminal connector.
(c) Improper soldering of cable (lead) with cable adopter.
This may result in overheating or welding of the terminal. In
some cases, melting of soldered joints is observed, resulting even in
falling of the lead. It is therefore recommended that brazing be re-
sorted to in place of soldering.
25.8 Dos and Do nots for HV Condenser Bushings
Dos
(a) DO check the packing externally for possible transit damage

before unpacking.
(b) DO unpack with care to avoid any direct blow on the bushing
or porcelain insulator.
(c) DO store the bushing in a shed or covered with tarpaulin to
protect it from moisture and rains.
(d) DO handle the bushing with manila rope slings without any
undue force on the porcelain insulator.
(e) DO clean the porcelain insulator thoroughly before taking

any measurement or mounting the bushing on the trans-
former. Check also for any foreign body adhered to the bottom
oil end porcelain.
(f) DO check the oil level by making the bushing vertical.
(g) DO check for leakage of oil from any of the joints. Each
bushing is tested with the oil immersed in the oil tank, and
so some traces of oil can be found, which actually is not
leakage.
(h) DO check the tan delta and capacitance on mounted bushing
with the jumper connection removed.
(i) DO maintain a log book of records of periodical checks as
mentioned the in the supplier’s O and M manual.
25.8.2 Do Nots
(a) DO NOT unpack the bushing from the crate unless required
to be mounted on the transformer.
(b) DO NOT use metal slings on porcelain and avoid undue jerks
while handling.
(c) DO NOT store the bushing outdoors without any protective
covering.
(d) DO NOT measure IR value or tan delta value without
thoroughly cleaning the porcelain.
(e) DO NOT dismantle or attempt to repair the bushing without
prior permission of the manufacturer.
(f) DO NOT fill the oil in the bushing without specific instruc-
tions from manufacturer.
(g) DO NOT climb the porcelain to tighten the top terminal. Use
an elevator or separate ladder for this purpose.
C HAPTER 26
Computerisation—A Tool to
Enhance Engineering
Productivity
R. Mitra
DGM/TRE
Transformers are custom designed products. Practically every order

requires a separate design and a new set of drawings. Though cer-
tain assemblies/components have been standardised, these consti-
tute a small part of the total design effort. A typical transformer may
consist of as many as 3000 components/sub-assemblies, and the com-
plete Engineering Information may require 500 Drawings/Bill of
Material. The effect of this huge design/drawing work is that the
issue of Engineering Information takes as much as 25% of the total
delivery cycle of the transformer.
All the components/sub-assemblies/assemblies that go into a
transformer have to be so designed that they can be assembled
properly, any mismatch identified at the time of manufacture is
difficult to rectify. In addition, all modifications carry a certain
penalty in terms of the cost of the product as well as the time taken
to manufacture. Since a large number of drawings have to be made,
they are made simultaneously, adding to the chances of mismatch.
Any Engineering Drawing has, basically, two sets of dimensions
that are the most important part of the drawing. These are
(a) dimensions which are independent variables, i.e. they are
derived from the specification, rating, customers require-
ment, etc.
(b) dimensions which are dependent, i.e. they are derived from
either the independent dimensions—based on certain
formulae, or are derived from certain standard practices or
design rules.
Using this logic, it is possible, if the independent dimensions are

available, to work out the dependent dimensions, and thus all the
dimensions needed to make the drawing can be worked out.
If we take a top down view of the drawing process, we will see that
an assembly consists of a number of components. Certain dimensions
derived from the assembly are common to a number of components.
When the drawings are being made, any mistake in the common
dimensions, either for the assembly or the component drawing
would create a mismatch. In the ordinary course, all drawings with
common dimensions should be checked at the same time to ensure
correctness. The fact that all such drawings may not be available, as
well as the possibility of human error creates mismatches that have
to be rectified later at a cost to the organisation.
A new approach has been taken to minimise such errors. Computer
programs have been developed, which take the dimensions identified
as independent variables as inputs. The logic for calculating the
dependent variables is built into the program, and the dependent
variables calculated. Based on these dimensions, and the relations
between them, a 3-D model of the component/assembly is generated
by the program. Finally, the program also generates the 2-D Produc-
tion Drawings. Since separate programs would cover different
assemblies, all the common dimensions are available while making
the component drawings. The chances of errors/mismatches are
thus reduced. Since all the data required to make the Bill of Material
are available, the program generates the Bill of Material also. The
advantages of such a system are:
(a) By giving one set of data for an assembly, a number of
drawings and the Bill of Material would be generated. Thus
the time taken to make the drawings and the Bill of Material
is reduced.
(b) Since the common dimensions would be given only once,
errors due to mismatch is reduced.
(c) As a 3-D model would be available, the designers can rotate
the model and visualise it from different angles and thus
make improvements.
(d) Fouling between different components/assemblies, excess
clearances would be visible on the computer screen in 3-D,
allowing corrections to be made before the drawings are
released, or components procured.
Computerisation—A Tool to Enhance Engineering Productivity 579
(e) The Bill of Material would be generated by the program,

and can be stored in a database for use by the planning
groups.
(f) Since all calculations are built into the program, there would
be commonality of logic. Mistakes due to adopting the wrong
formulae by an individual are therefore eliminated.
(g) Any changes in the design philosophy would entail modifica-
tions in the program and the need to ensure proper changes
in the individual designs would be eliminated. This would
ensure better design control.
Based on the above philosophy a Parametric Drawing Genera-
tion System (SYPAD) has been developed. The system is suitable
for the design of transformers following BHEL’s design methods.
The system is very comprehensive and would generate the drawings
and the bill of material for a large variety of transformers with
different ratings.
The system has been designed with menu driven prompts so that
an user without any knowledge of 3-D modelling software or
database packages can use the system. The only need is that the
user knows how to use AutoCAD, which is a common CAD software.
The basic modelling software used in the system is M/s Autodesk’s
Mechanical Desktop. The menu screens for making the system user
friendly as well as the calculation etc has been developed in Visual
Basic, while Oracle is the backend database.
As an example of the increase in productivity, let us take the case
of the model shown in Fig. 26.1(f). To make the complete assembly
seven drawings of the individual components would have to be made.
All seven drawings and the Bill of Material can be made through one
program by giving 40 independent variables as input. The program
calculates, using various formulae, about 160 different dimensions
relevant to the seven drawings. One model of the assembly, and
seven models of the individual components are generated and the
2-D Production Drawings and Bill of Material are automatically
generated. The Billl of Material data is also stored in a database for
future use. Normally, about three weeks would have been required
to make all the drawings. Using the Parametric Drawing System,
the same seven drawings can be made in two days.
(a) Rectangular Turret

(Double Bushing)
(f) Core and Endframe Assy
(b) Rectangular Turret
(g) Header Pipe Work

(c) Top Tank
(h) Cooler Pipe Work

(d) Bottom Tank
(e) Limb Core with Coil Assembly (i) Inside Mounting

(Top and Bottom Yoke Shunt Assy)
Figure 26.1 Shows some of the assemblies which have been

developed through the system.
C HAPTER 27
Condition Monitoring,
Residual Life Assessment and
Refurbishment of Transformers
C.M. Shrivastava
T.S.R. Murthy
The need for reliable and stable systems is being increasingly felt.
Thus, emphasis is now being laid on 100% capacity utilisation and
the availability of equipment for reliable operation of the system.
Fortunately, many tools are now available or underdevelopment,
which can be used for condition monitoring of the transformer.
With a view to cover detailed information about DGA and Condi-
tion Monitoring, viz residual life assessment, this chapter has been
divided into two sections:
Section A: Dissolved Gas Analysis Interpretation.
Section B: RLA and Refurbishment.
SECTION A
Dissolved Gas Analysis Interpretation
Incipient faults in oil filled transformers are usually the result of
electrical or thermal stresses in either the transformer oil or
insulating materials.
It is known that such excessive stresses produce a mixture of
gases characterstics of which give an indication of the type of fault
and location associated with the fault.
It is recommended that analysis of dissolved gases in transformer
oil by gas chromatograph equipments is made at the time of commis-
sioning and then every six months for transformers of 145 kV class
and above.
27.1 Analysis Method
27.1.1 Sampling
Oil in transformers can be sampled through the drain or sampling

valve near the bottom of the tank. Special care is to be taken not to
introduce air, foreign matter, or dirty oil into the sampling con-
tainer. For this purpose, first 0.5–1.0 litre of oil from the transformer
is to be over-flown through the oil container. The oil sample must
not be exposed to air before analysis.
27.1.2 Analysis of Fault Gases
The gases to be analyzed and the criteria for the gases found in the
transformer oil are tabulated in Table 27.1.
Table 27.1 Gases to be Analyzed and Criteria
Sl. No. Operating condition Gases to be analysed

1. Normal O2, N2, H2, CO, CO2, CH4
2. Abnormal H2, CH4, C2H2, C2H4, C2H6
3. Deterioration CO, CO2, CH4
The generation of gases in oil by some typical faults in transformer

active part models are shown in Table 27.2.
Table 27.2 Gas Content in Oil by Faults
Sl. No. Type of faults Decomposable gases in transformer oil

1. Arcing in oil CH4, C2H4, H2, (C2H6, C2H2, C3H6, C3H8)
2. Overheating of solid
insulating materials CO, CO2, (H2, C2H4)
3. Overheating of oil and
paper combination CH4, C2H4, CO, CO2, H2
4. Arcing of oil and paper
combination H2,C2H2, CO,CO2, (C2H4)
Note. ( ) shows gas contents which appear rarely.
Condition Monitoring, Residual Life Assessment 583
To identify the nature of fault from the DGA results, the ratio
technique of IS/IEC is used as per Table 27.3.
Table 27.3 DGA Findings and Inspection
Sl. no. Characteristics Code of range ratios Findings

C2H2 CH4 C2H4
C2H4 H2 C2H6
1. <0.1 0 1 0
2. 0.1 to 1 1 0 0
3. >3 2 2 2
4. No. Fault 0 0 0 Normal ageing
5. Partial discharges 0 1 0 Discharges in gas filled
of low energy cavities resulting from
density incomplete
impregnation or super
saturation or cavities or
high humidity.
6. Partial discharges 1 1 0 As above, but leading
of high energy to tracking of perfora-
density tion of solid insulation.
7. Discharges of low 1Æ2 0 1 Æ 2 Continuous sparking in
energy oil between lead connec-
tion of different potential
or to floating potential.
Breakdown of oil between
solid materials.
8. Discharges of high 1 0 2 Discharges with power
energy flow through arcing.
Breakdown of oil between
winding or coils to earth.
Selector breaking current.
9. Thermal fault of 0 0 1 General insulated conduc
low temperature -tor overheating.
<150 °C
10. Thermal fault of 0 0 2 Local overheating of core
low temperature due to concentration of
range 150 °C to flux. Increasing hot spot
300 °C temperature varying from
small hot spots in core,
shorting links in core,
(Contd.)
Table 27.3 (Contd.)
Sl. no. Characteristics Code of range ratios Findings

C2H2 CH4 C2H4
C2H4 H2 C2H6
overheating of copper
due to eddy currents, bad
contacts/joints (pyrolitic
and carbon formation) up
to core and tank circula-
ting current.
11. Thermal fault of 0 2 1 -do-
medium tempera-
ture range (300° C
to 700° C)
12. Thermal fault of 0 2 2 -do-
high temperature
range >700° C
Depending upon the fault gases ratio obtained in the DGA analy-
sis, the intensity of inspection is established. It is recommended that
incase of minor change in DGA, results, at least three readings
(tested in the same lab) are to be analyzed to arrive on better conclu-
sion.
27.2 Physical Inspection
Before taking total outage of transformer following preliminary

physical inspection, tests should be carried out to establish the rea-
son for the increasing trend of fault gases.
27.2.1 Physical Preliminary Inspection
Sl. Description Status

No. OK Not OK
1. Check for any hot spot developing with the

help of thermovision camera.
2. Monitor temperature of tank, rim bolts and
turret bolts.
(Contd.)

No. OK Not OK
3. Conduct PD Measurement by acoustic meter.
4. Check core isolation with yoke clamp and tank.
5. Check BDV and PPM of transformer oil.
6. Check bushing terminals (HV and LV) for
any looseness/heating mark etc.
7. Study the temperature log with relation to
loading of transformer.
8. Check cooling equipment like oil pumps and
fans for proper flow of oil and air.
9. Check valve circuitry of cooler pipe work and
radiator etc. for proper opening.
10. Conduct FRA test if previous benchmark
data is available for comparison.
In the second stage for finding the root cause and doing remedial
rectification, the following points should be checked through inspec-
tion covers after draining out oil from the transformer.
27.2.2 Core

No. OK Not OK
1. Carry out general inspection inside the tank
to see any abnormal heating/hot spot etc.
2. Check for connection and single point
earthing of top and bottom yoke shunt.
3. Check proper earthing of wall shunt.
4. Check for proper connection of earth shield
provided on end legs, yoke etc.
5. Core isolation between:
(a) Clamp to core
(b) Clamp to tank
(c) Core to tank
(d) Yoke bolt to core
(e) Yoke bolt to clamp
6. Check isolation between end frame to end
frame (400 kV transformers only)
7. Check condition of cross beam and end plate
area, tapchanger support and cover locking etc.
27.3.3 Terminal Gear

No. OK Not OK
1. Check HV/LV line lead connection for heating
and burning marks.
2. Check bus bar to bus bar joint for proper
tightening and brazing.
3. Check for proper tightness of LV flexible
jumpers.
4. Check all tapping leads connected on tap-
changer for proper tightness and heating
impression.
5. Check earthing connection of shield provided
on turrets.
6. Check all barriers for heating and black
soothing mark if any.
27.2.4 Winding

No. OK Not OK
1. Check lead connection after removing paper
insulation for proper metalised shielding and
brazed joint.
2. Check line lead duct for black soothing mark/
carbonization.
3. Check general condition of barriers provided
on winding.
27.2.5 Bushings

No. OK Not OK
1. Check oil level in bushing conservator.
2. Check test tap connection and earthing.
3. Carry out physical inspection of oil end
of bushing.
4. Check proper tightness of terminal provided
on top of bushing.
5 Check any tracking mark on oil end of bushing.
6. Check condition of corona shield and fitting.
27.2.6 Tap Changer

No. OK Not OK
1. Carry out physical inspection of selector
terminal connection.
2. Check proper tightness of leads connected
on selector.
3. Check condition of fixed and moving contacts
of selector for arcing/carbonization.
4. Check for proper alignment of selector assembly.
5. Check condition of oil filled in diverter tank.
6. Check for any leakages through diverter tank
to main tank.
7. Check for any damage in diverter switch assembly.
8. Check the continuity of diverter transition resistor.
9. Check healthiness of copper braid connection.
27.2.7 Insulating Oil

No. OK Not OK
1. Check BDV and PPM of oil.
2. Check tan delta and resistivity of oil at 90°C.
Note: (a) Internal inspection should be carried out in presence of the
manufacturer's representative.
(b) It is possible that some of the checks as listed above may not be
assessible without untanking the transformer. All such activity points
are to be checked after untanking, and if required the coil assembly is
to be removed. All above activities shall be carried out strictly under
supervision of the manufacturer's representative only.
SECTION B
RLA and Refurbishment
27.3 Residual Life Assessment (RLA)
Presently, most of the electrical utilities carry out Dissolved Gas

Analysis (DGA) of the transformer oil, apart from routine electrical,
physical and chemical tests on oil to determine the health of the
transformer. However, the DGA gives indication only regarding the
incipient fault related to corona, arcing, or overheating that takes
place in the transformer. It does not give any conclusive evidence
related to the degradation of cellulose paper, and thereby does not
give information regarding the health of the transformer winding
or remaining life of the transformer. Therefore the Residual Life
Assessment (RLA) study is carried out to predict the health of the
transformer insulation and remaining life of the transformer.
It is difficult to assess the life of the transformer because of
the complex behaviour of insulation. The ageing of insulation in
transformers is influenced by short term and long term over loads,
number and intensity of short-circuits, incidence of lightning, and
internal faults.
The ageing behaviour is likely to be different for different types
of transformers. The life span of the transformer, thus, depends
initially on the design, and quality of manufacture, and later on
service conditions and maintenance standards.
These factors vary considerably and affect the useful span of
service life, which therefore needs to be taken into account for
residual life assessment. The recent development of various
techniques for detecting the incipient fault conditions have improved
to some extent, the life expectancy of transformers, by resorting to
corrective actions in time.
During the natural ageing of transformers, the insulation of wind-
ing, which is cellulose paper, deteriorates. Such deterioration can be
well assessed by subjecting the cellulose paper to various tests. But
the windings of oil filled transformers are generally inaccessible.
However, by draining out the oil, an inspection of the outer layer
is possible though it is tedious and cumbersome. It is necessary
to assess the condition of the cellulose insulation at the hot spot
developed within the winding, which is inaccessible. An easy method

of finding the integrity of winding insulation is by chemical analysis
of the oil in which the winding is immersed.
Cellulose insulation degrades due to heating or electrical
breakdown, resulting in the production of furfural derivatives, which
dissolve in oil. Hence, the chemical analysis of the transformer
oil gives evidence of changes that are taking place in the winding
insulation during operation.
The constituents of dissolved gases in oil are utilized in the early
detection of incipient fault within the transformer. From the ratios
of particular pairs of gases, the fault conditions are diagnosed. It is
required to keep a record of gas analysis data from the date of instal-
lation, and provide the same for the assessment. In addition to the
above, it is recommended to keep conventional test data and the
history of transformer in the format provided as per Annexure-I.
An analytical technique has been developed to ascertain the re-
maining life of transformer winding cellulose-based insulation paper.
The technique involves extraction of chemical compounds from
transformer oil, its quantitive estimation using High Performance
Liquid Chromatography (HPLC), conversion to degree of polymer-
ization, and computation of the life. In addition to the above, fault
gases are analyzed using expert system software, along with electri-
cal and physiochemical properties of transformer oil monitored from
the date of commissioning and service life.
During the life assessment studies, the transformer oil samples
drawn are analyzed for all conventional properties in addition to spe-
cial properties. These inputs are then compared with the data bank
available. Depending on the trends observed, recommendations are
made about the status of the transformer in terms of residual life
and/or need for replacement/reclamation of oil, and frequency of the
transformer oil analysis.
27.4 Different Techniques for Life Estimation
Different techniques have been developed to estimate the life of

transformer winding insulation. Some of the methods are mere
extensions of conventional analytical techniques in terms of data
interpretation. The techniques generally followed are summarized
below:
27.4.1 Estimation from CO2 and CO
Dissolved Gas Analysis (DGA) is the most important parameter,

which helps in identifying the faults inside the transformer. From
the quantities of CO and CO2, the status of paper insulation is
estimated by floating ARV MAP graphs. These ratios are to be
supplemented with the degree or polymerisation of furfural values
for estimation of life.
27.4.2 Estimation from Furfural
Cellulose insulation degrades due to heating or electrical break-

down, resulting in the formation of derivatives of furfurals which are
soluble in oil. However, 2-furfural dehyde is the major constituent
of the liberated furfural derivatives. By monitoring the levels of
2-furfural dehyde by High Performance Liquid Chromatography
(HPLC), and comparing those values with the data bank generated
for the transformers of various ratings and age, the remaining life
can be assessed. The method has a limitation in terms of solubility of
furfural dehyde in transformer oil, viz. its generation, and has to be
complemented by other techniques.
Oil Sample Collection

The oil sample is generally taken from the bottom valve of the trans-
former. Depending on the oil quality and furfural content, it can be
decided whether collection of further oil samples for ascertaining
the status of transformer are required. Approximately five litres of
the oil are to be collected in stainless steel bottles/glass bottles by
following relevant standard procedures. Oil sample may be required
a minimum three times of over a period of 6–8 months time during
the transformer in service, to ascertain the rate of rise of furfural.
Paper Sample Collection
The paper sample must be collected from the transformer after shut
down from lead insulation of HV and LV side after opening the trans-
former and lowering the oil level. Also, retaping of the paper with
fresh paper on the removed portion must be done. The size of paper
required for testing of DP is approximately 30 cm2, weighing around
400 mg to 500 mg.
Compilation and Issue of Test Report

The results of samples collected can be analyzed, indicating the re-
sidual life of the transformer.
27.4.3 Estimation from Degree of Polymerization
The Degree of Polymerization (DP) is a direct parameter, which has

been established as a linearly varying identity under thermal stress.
Hence, a measure of DP of winding insulation paper can give the
remaining life. However, such paper samples are inaccessible. DP
values of paper samples from the next available location, that is lead
insulation, are measured. Hence, the analytical data from this
method has to be interpreted along with furfural values measured
from the transformer oil.
27.5 Methodology Adopted
The remaining life of the transformer can be estimated by measur-

ing the furfural derivative from oil and the degree of polymerization
(DP) of cellulose paper from the lead insulation of windings. The life
of the transformer is basically the life of the winding insulation.
Normal and abnormal stress on winding results in over heating of
the oil or winding insulation paper. Hence, levels of deterioration
are monitored by the estimation of furfural from transformer oil,
and the degree of polymerization (DP) from lead insulation paper.
The High Performance Liquid Chromatography (HPLC) tech-
nique is employed for estimation of furfural content after extraction
from the transformer oil sample with a suitable solvent. Since it is
the degree of polymerization (DP) which determines the age of the
transformer, the furfural content measured is corroborated with the
theoretical value of furfural calculated from DP measured for a few
winding paper samples of the same transformer. The degree of poly-
merization is measured for the cellulose paper sample, preferably
drawn from all HV and LV leads.
The method involves synthesis of a special solvent, dissolution of
paper in solvent, measurement of viscosity, and its conversion to
DP. The DP value thus computed is compared with the DP from
furfural estimation.
In addition to the above, special tests on the transformer oil

samples drawn are conducted for dissolved gas analysis (DGA) and
the conventional test for physiochemical and electrical properties as
indicated below. The data generated can be logged on to the expert
system diagnostic program to ascertain the transformer condition in
terms of stresses it has undergone during service.
Final recommendations can then be made, based on the above
analysis regarding the residual life and reconditioning/ replacement
requirements.
27.6 Conventional and Special Tests on Oil Samples
27.6.1 Physical Tests
(a) Specific gravity ASTMD 1298–85 (IS 1448-P-16) 1967

(b) Refractive index ASTMD 117–89
(c) Interfacial tension ASTMD 971–91 (IS 6104–1971)
(d) Kinematic viscosity – IS 1448P–25 1960
(e) Flash point IS 1448P–10 1971
27.6.2 Chemical Tests
(a) Total acidity ASTMD 974–92

(b) Sludge content ASTMD 1698–84
(c) Moisture content ASTMD 1533–88
27.6.3 Electrical Tests
(a) Tan delta ASTMD 924–92

(b) Resistivity
(c) BDV ASTMD 1816–84 a.
27.6.4 Special Tests
(a) Rotary Bomb Oxidation test RBOT ASTMD-2212.

(b) Elemental analysis
(c) Dissolved gas analysis (DGA) ASTMD 361–93
(d) DGA Interpretation – IEC 60599/IS 10593/IEEE Std.62–1995,

ROGERS RATIO METHOD.
(e) Furfural content – IEC 61198.
(f) Degree of plymerisation (DP) test—to be done on availability
of paper sample as per IEC 60450 and calculation of furfural
content in oil from DP value.
27.7 Refurbishment
The transformer is to be withdrawn from service as per planned shut

down. For refurbishment, the core and coil assembly is taken out of
the tank. The time required for refurbishment (servicing/reinstall-
ing and commissioning) of the transformer can add upto to about
four weeks.
27.7.1 Activities Involved
l Low voltage tests i.e.

(i) Winding resistance.
(ii) Magnetizing current
(iii) Magnetic balance.
(iv) Ratio.
(v) Insulation resistance.
(vi) BDV of oil.
l Draining of complete oil.
l Removal of bushings, turrets and pipe work.
l Shifting of transformer to service bay nearby where servicing is
planned.
l Removal of tank cover.
l Lifting of core and coil assembly.
l Cleaning and removal of deposit, dirt and sludge.
l Checking of yoke bolt and core insulation.
l Tightening of terminal gear and core.
l Servicing of tapchanger and replacement of defective compo-
nents.
l Retanking of core and coil assembly in tank and refitting of cover.
l Vacuum pulling and dry out.
l Rewrite as Vacuum pulling and dry outs.

l Shifting of tank to plinth.
l Assembly of turrets and bushings and other accessories.
l Oil filling, pressure testing for ensuring that there is no oil
leakage, and final dry out by hot oil circulation.
l Final painting of transformer and accessories.
l Pre-commissioning checks.
27.7.2 Facilities Required
l Service bay with mobile crane or mobile covered shed,

fabricated with scaffolding pipes and GI sheets, tarpolene, rails,
rollers, etc.
l Steel tray made of sheet.
l N2 gas of 99.9% purity and dry –50ºC dew point.
l Industrial concealed heaters.
l Vacuum pumps sufficient to maintain vacuum in the tank up to
0.5 torr.
l Oil filter machine, high vacuum 4500/6000 lit/hr.
l Oil storage tank to accommodate the total quantity of
transformer oil.
l Mobile crane suitable to lift the core and coil assembly from tank.
l Platform/ladders.
l Hand tools, electrician tools.
l Testing equipment for dielectric test, water content, ratio,
resistance of winding, megger, multimeters, etc.
l Skilled and unskilled man power.
27.7.3 Materials Required for Opening

and Overhauling the Transformer
l One set of complete erection gasket.

l All manufacturing gaskets related to tank assembly.
l Any other item found defective to be replaced.
27.7.4 Advantages of Refurbishment
l Pre-determination of incipient pre-matured fault (if any).

l Improvement in cooling due to removal of deposit, dirt and

sludge.
l Ensuring tightness of all bolted joints.
l Reduction in noise and vibration after tightness of core clamp
and bolts.
l Re-conditioning of insulating oil.
Many transformers designed and manufactured 15 to 20 years back
may not be able to perform satisfactorily because of the auxiliaries,
fittings, and terminal equipments with old technologies. To improve
their performance, it is possible to replace these vital fittings with
fittings of the latest technology. Important fittings which can be
changed during refurbishment are:
l Bushings
l Tapchangers
l Buchholz relay
l Oil conservator with Atmoseal
l Oil temperature indicators
l Winding temperature indicators
l Removal of tertiary winding: Tertiary winding is one of the
weak parts in transformers. Many transformers have failed due
to tertiary. As per latest CBIP recommendations, it is not
mandatory to provide tertiary winding for transformers upto
100 MVA. This will not only help increase the reliability of the
transformer, but will also provide extra space for uprating of
transformer capacity.
l Redesign windings: The windings are a vital part of the transfor-
mer. With the advent of latest type of windings, like interleaved
disc and partially interleaved disc windings, improved quality
of insulating materials and transformer oil, and computer
aided design techniques for analysis of the transformer under
transient and power frequency conditions, it is now possible to
optimize the winding design and provide 10% to 50% more
capacity in the same window as available for old transformers.
27.7.5 Methodology Proposed for Refurbishment
l Use the existing core, tank and fittings. These are reusable
without having an adverse effect on the performance under
uprated condition. These cover about 50%–60% of the cost of

the transformers.
l Redesign the windings with the available computer aided
tools/technology so that it is able to meet the latest IEC
requirement withstand capability. Modify the clamping
arrangement to increase the short-circuit power frequency
conditions strength.
l On the existing tank, provide extra cooling arrangement to
meet the dissipation of extra losses due to increased capacity,
and meet the oil and winding temperature rises within limit.
Considering the high return of scrapped copper winding, the
total cost of uprating is about 40–45% of the cost of new
transformers of the same rating.
27.8 Conclusion
Due to paucity of funds, it is not practical and economical to replace

old units with new one. In such cases, refurbishing/retrofitting is
an economical and viable alternative. The main advantage of refur-
bishment on site is that there is no need for rail or road transport.
Consequently, the transportation cost and risks are eliminated by
RLA and refurbishment. The transformer life is extended, and the
transformer made suitable for providing satisfactory services/
performance. Uprating of the transformer alongwith refurbishment
helps the user in meeting system demand uninterruptedly.
ANNEXURE I
Format for Data Collection History of Transformer
1. Type of transformer :
2. Rating :
3. Date of manufacture :
4. Sl. No. and make :
5. Date of commissioning and :
location
6. Insulation details :
(a) Weight of oil
(b) Weight of paper
7. Type of winding used (tick) : Spiral/Cont.Disc.
8. Type of cooling :
9. Make of oil used :
10. Quantity of oil used :
(a) In tank
(b) In conservator
11. Date and number of topping
ups done :
12. Quantity of oil used for
each topping and last date
of topping :
13. Details of breakdowns (minor) :
14. Details of breakdowns (major) :
15. Temperature data :
(a) Oil temp.
(b) Winding temp.
16. DGA and other data
H2, CO, CO2, CH4, C2H6, C2H4, C2H2
Oil parameters
Breakdown voltage (kV) :
Moisture (PPM) :
Resistivity X 1012 (ohm.cm) :
Tan delta (90°C) :
Total acidity (mg/g) :
IFT at 27 °C (N/m) :
Flash point (°C) :
Furfural content :
17. Remarks for any follow up :
Signature:
Name:
Designation:
Date:
C HAPTER 28
Transformers: An Overview
S.N. Roy
P.T. Deo
In the previous chapters, various technological aspects of transform-

ers, starting from core design to final testing and commissioning
at site is explained in detail. Though a transformer is a simple and
static electrical equipment, its outage can cause interruption to all
spheres of daily life.
Specification is the basic document on the basis of which the
design is finalised, and final product is realised. To begin with, the
user of the machine has to study all site condition requirements and
spell out the critical parameters. The basic data requirement is
defined in various chapters and has to be strictly followed. On
the basis of this basic requirement, the design of a transformer is
conceptualised. In this era of computers, most of the design is com-
pleted by feeding the input data. Since the programs are already in
use, the logics behind these programs are not realised by the
new engineers. It is necessary that a fresh engineer must study the
formulae in detail, and try to understand the logics which have been
used in the programs.
Next is the incoming material for which different manufacturers
are already available in the market. The need for vendor development
is true today from a quality point of view as well as prompt delivery.
In an organisation, methods and tools are available to locate a vendor
and then evaluate the work from a technical competence and quality
point of view. The raw material has to be arranged in a way that the
different assemblies and sub-assemblies are completed for realising
a final product. The major activities which constitute formation of a
transformer are core, winding, tank, tapchanger, bushing, cooling
equipment, etc. So, for all such sub-groups, one has to look for raw
Transformers: An Overview 599
material, working platform/machines, skilled manpower, etc. Each

activity has to be critically planned and monitored at every stage.
Producing a quality transformer is the need of the day and the
factors which can give reliable service can be summarised as under:
(a) The specification should cover the system requirement and
any other non-standard duty-performance that the trans-
former has to see.
(b) The designers should make use of all combinations of duty
cycle requirements and put these into a computer program to
generate the most economical design. The facility available
with the manufacturer for producing core, winding, assembly,
and processing must be taken into account. Otherwise, the
design finalised may not be possible to handle in the shop
floor.
(c) The selection of components like bushing, tapchanger,
protective devices etc. must take care of the over-voltage,
transient overloading and other such conditions.
(d) The raw material and component sourcing must be done from
limited vendors which meet the specification requirement
in totality. To meet the short delivery of transformers,
the green-channel vendor development and supply may
be considered. This means that the vendor is capable of
producing in accordance with the job specifications and
quality assurance procedures, and hence quality control
checks at every stage can be withdrawn.
(e) A well trained work force is necessary to understand the
complex drawings, and produce the equipment. A slight
deviation unnoticed shall make the transformer susceptible
to failure at the test bed or in service prematurely.
(f) Cleanliness in the shop floor particularly insulation, winding,
assembly, etc. must be maintained. Dirt particles, which can
also be conductive, are extremely harmful for transformers,
which are normally high voltage equipment.
(g) The insulation items used in transformer winding, core,
assembly, etc. are mostly hygroscopic in nature. Processing of
raw material before starting manufacture of insulation items
can give a very stable dimension to these items. The items
used directly within the winding play an important role for
dimensional stability. Any difference in the dimension of

windings other than as specified by the designer, can produce
extremely high amount of short-circuit forces.
(h) System should prevail in every sphere of design, production,
testing, etc. Every design rule, operation instruction, etc.
must be documented, and made available at the work place.
(i) There should be a system for feedback from the shop floor,
testing, and the customer, and records maintained in a well
formatted data bank format. Such information helps in
re-looking into the design for any corrective measures and
improvements from a service reliability point of view.
For transformers with higher ratings with respect to MVA and
kV, even a minor deviation if not detected at the processing stage
may create problems during final assembly and testing. The process
can be either design, raw material, manufacturing, assembly, or
testing. It is imperative to study all these processes into the minut-
est detail. We call such study of the processes as “Process Mapping”.
Under this head, each activity is critically examined. The areas can
be defined as below:
(a) Specification and Design.
(b) Manufacturing.
(i) Core
(ii) Winding
(iii) Assembly
(iv) Processing
(v) Testing and despatch
(c) Site assembly, pre-commissioning.
(d) Loading to the capacity and feedback.
The specification which defines the basic requirement of the
equipment and site condition is the tool to start a design. Knowing
the capitalisation rates of no-load loss, load-loss, and auxiliary loss,
the initial design work starts. One is supposed to know the transport
limitation with respect to dimensions and the weight which can be
handled. A compromise is made with respect to the material cost and
loss-capitalised cost, and the design is finalised. A low loss design
shall need more weight of material, i.e. core, copper, oil, etc. But the
Transformers: An Overview 601
transport restrictions imposed may cause a re-look into the design

and make a final feasible design from all aspects. One can look into
the strengthening of the small road bridges etc. from transport point
of view. A thorough route survey is undertaken in such a case.
The manufacturing process for the sub-groups as defined in (b) can
be looked into in detail. The data-bank and feedback of the shop floor
for a minimum of 24 months becomes a tool for taking-up the correc-
tive measures. The non-conformities are tabulated, and the critical
failures identified. A group can be formed with design, production,
technology, and quality-control, to review these non-conformities.
The reason is identified and the corrections are put into the new
technology/process specification. For success, a process-owner can
be identified. He can take the lead role in corrective measures and
final implementation.
After final testing at works, the site assembly and commissioning
becomes quite important from an operation point of view. Only
qualified technicians should be allowed to handle the transformer at
the site. Most of the transformer manufacturers have their field
check-points. During assembly, all such checks should be verified
and recorded in a format. Any deviation must be critically examined,
and referred to the manufacturer for final acceptance or corrective
action. The check-list jointly signed by the supplier and the owner
shall form as the base document. This can be maintained as a history-
card. After final commissioning, loading to the required capacity is
done. In case of any interruption or maloperation of the equipment,
the same must be immediately referred to the manufacturer. These
incidents can be recorded in the history-card, which shall give the
overall view of this vital equipment.
For easy access to the manufacturer and to seek urgent attention,
it is now possible to establish a centralised contact station (website),
and all the users can be connected directly on line. Even the users at
remote areas can make use of the latest advances of information
technology to get prompt services of the manufacturer.
The transformer industry is now more than 100 years old. We at
BHEL have produced more than 3700 transformers, contributing
approximately 230000 MVA to the power system in India and abroad.
Producing these units is a craftsman’s job since every stage of manu-
facture is manually accomplished. As such, it needs a highly skilled,
trained, and motivated work force to achieve the desired quality. In
the present day scenario, it is possible to identify the technical

institutions that can be involved with the training in different areas
of manufacturing, i.e. core, insulation, winding and assembly, etc.
This will help in generating qualified and trained manpower for the
industry.
Finally, it is always possible to simplify the processes of the
manufacturing stages rather than training our people for the
complex processes generated by us. It is necessary to think in the
direction of making our activities simpler. Maybe we can adopt the
modular concept of design standardisation and manufacture of the
various sub-assemblies so that mistakes are minimised, and we
produce a reliable piece of equipment for years to come.
C HAPTER 29
Solved Examples
1. A 60 MVA, 11/132 kV three-phase, 50 Hz, generator

transformer has guaranteed parameters as, load loss 150 kW
at full load at 75°C, no load loss 30 kW and percentage
reactance 9% at rated load. Calculate
(a) percentage regulation at
(i) full load and unity power factor
(ii) 3/4 load and 0.85 lagging p.f.
(b) Efficiency at
(i) full load and unity p.f.
(ii) 3/4 load and 0.85 lagging p.f.
(c) Maximum efficiency and load at which it occurs.
Solution
(a) (i) r = (150 ¥ 100)/(60 ¥ 1000) = 0.25%, x = 9%, cos q = 1.0,
sin q = 0.0
Percentage regulation = r cos q + x sin q + (x cos q –
r sin q)2/200
= 0.25 ¥ 1 + 0.0 + (9 ¥ 1 – 0.25 ¥ 0.0)2/200
= 0.655%
(ii) r = {150 ¥ (0.75)2 ¥ 100}/(60 ¥ 0.75 ¥ 1000) = 0.1875%,
x = 6.75%
cos q = 0.85, sin q = 0.53
Percentage regulation = 0.1875 ¥ 0.85 + 6.75 ¥ 0.53 +
(6.75 ¥ 0.85 – 0.1875 ¥ .53)2/200
= 3.9%
(b) (i) Percentage efficiency = (output ¥ 100)/(output + Cu
loss + iron loss)
= (60 ¥ 1000 ¥ 100)/(60 ¥ 1000 + 150 + 30)
= 99.70%
(ii) Percentage efficiency = (60 ¥ 1000 ¥ 0.75 ¥ 0.85 ¥ 100)/

{60 ¥ 1000 ¥ 0.75 ¥ 0.85 + 150 ¥ (0.75)2 + 30}
= 99.70%
(c) Maximum efficiency at unity p.f. at load = F.L. ¥ ÷(iron
loss)/(F.L. Cu loss)
= 60 ¥ ÷(30/150) = 26.83 MVA
Maximum efficiency occurs when Cu loss = Iron loss
hence, max. efficiency = (26.83 ¥ 1000 ¥ 100)/(26.83 ¥ 1000
+ 30 + 30)
= 99.78%
2. A 100 MVA, 220/66 kV, Y/Y, three-phase, 50 Hz transformer
has iron loss 54 kW. The maximum efficiency occurs at 60%
of full load. Find the efficiency of transformer at:
(a) full load and 0.8 lagging p.f.
(b) 3/4 load and unity p.f.
Solution
Cu loss at 60% of full load = 54 kW
Cu loss at full load = 54 ¥ (1/0.6)2 = 150 kW
(a) Percentage efficiency = (100 ¥ 1000 ¥ 0.8 ¥ 100)/(100 ¥ 1000
¥ 0.8 + 54 + 150)
= 99.75%
(b) Percentage efficiency = (100 ¥ 1000 ¥ 0.75 ¥ 1.0 ¥ 100)/{100
¥ 1000 ¥ 0.75 ¥ 1.0 + 54 + 150 ¥ (0.75)2}
= 99.81%
3. A 33/11 kV, 50 Hz, single-phase transformer has core loss of
10 kW and exciting current of 0.9 amp, when its HV side is
energized at rated voltage. Calculate the two components of
the exciting current.
Solution
Exciting current Io = 0.9 A, supply voltage V1 = 3300 V, Core loss
Pc = 10 kW
Hence core loss component Iw = Pc/V1
= (10 ¥ 1000)/33000
= 0.303 A
and magnetizing component Im = ÷(I2o – I2w)
= ÷(0.92 – 0.3032)
= 0.847 A
Solved Examples 605
4. Two transformers of same voltage ratio, rated at 315 MVA

each are connected in parallel to supply a load of 700 MVA
at 0.8 p.f. lagging. The per phase resistance and per phase
reactance of the first transformer are 2% and 11% respec-
tively, and of the second transformer are 1.5% and 12%
respectively. Calculate the load shared by each transformer.
Solution
Percentage impedance of first transformer Z1 = (2 + j11) % = 11.18
–79.69°
Percentage impedance of second transformer Z2 = (1.5 + j12)% =
12.09 –82.87°
Z1 + Z2 = (3.5 + j23)% = 23.26 – 81.35°
Load S = 700 –– cos–1 0.8 = 700 –– 36.9°
Load shared by first transformer S1 = S ¥ Z2/(Z1 + Z2)
= 700 – – 36.9° ¥ 12.09 – 82.87°/23.26 – 81.35°
= 363.84 – – 35.38°
= 363.84 MVA at 0.81 p.f. lagging
Load shared by second transformer S2 = S ¥ Z1/(Z1 + Z2)
= 700 –– 36.9° ¥ 11.18 – 79.69°/23.26 – 81.35°
= 336.45 –– 38.56°
= 336.45 MVA at 0.78 p.f. lagging
5. Calculate the percentage reactance of a 50 MVA, 33/220 kV,
50 Hz, 3-phase star/star connected transformer having the
following manufacturing information: Volts/turn-90, height
of HV and LV coil-2100 mm, radial depth of HV and LV coils-
80 mm and 50 mm respectively, mean length of HV and LV
turns-3.2 m and 2.2 m respectively and the radial gap between
the HV and LV coils-90 mm.
Solution
The percentage reactance X = 2pf ¥ m0 ¥ Lmt (AT)/(Lc ¥ Et) ¥ {a + (b1
+ b2)/3} ¥ 100
AT: current in HV = (50 ¥ 103)/(÷3 ¥ 220) = 131.2 A
turns in HV coil = (220 ¥ 103/÷3)/90 = 1411
hence AT = 131.2 ¥ 1411 = 185123,
Lmt = (3.2 + 2.2)/2 = 2.7 m
X = 2 ¥ p ¥ 50 ¥ 4p ¥ 10–7 ¥ 2.7 ¥ 185123/(2.1 ¥ 90) ¥
{0.09 + (0.08 + 0.05)/ 3} ¥ 100
= 13.92%
6. Calculate the air core reactance of an HV winding of 40 MVA,

21/11.5 kV, 3-phase, 50 Hz, delta/star connected transformer
having the following manufacturing details. HV coil: number
of turns-260, inside diameter of coil-940 mm, radial thickness
of coil-90 mm, and height of coil-1960 mm.
Solution
Air core reactance is calculated by the formula Xa = 2 p f Lsat
where Lsat = N2 ¥ Dwm ¥ K
where N = no. of turns in the coil
Dwm = inside diameter of coil
Factor K = 9.9 ¥ 10–7/(0.45 + a + r + 0.5 ¥ r)
where a = height of coil/inside diameter of coil
r = radial thickness of coil/inside diameter of coil
hence,
f = 50, N = 260, Dwm= 940, a = 1960/940 = 2.085,
r = 90/940 = 0.096
K = 3.70 ¥ 10–7
So, Xa = 2 ¥ p ¥ 50 ¥ 2602 ¥ 0.94 ¥ 3.7 ¥ 10–7
= 7.40 ohm
Percentage air core reactance = (7.4 ¥ 100)/{(21 ¥ 103)/634.92}
= 22.37%
7. Show the thermal ability to withstand short-circuit of dura-
tion 3 seconds of oil-immersed 50 MVA, 21/11.5 kV, 3-phase,
delta/star connected transformer having 12% impedance. The
Cu conductor cross-section area of HV and LV coils are 350
mm2 and 1050 mm2 respectively. The allowed temperature
rise of the windings at rated condition is 55º C over ambient
temperature of 50º C. The short-circuit apparent power of the
HV system is 9500 MVA and of the LV system is 500 MVA.
Solution
Calculation method for thermal ability to withstand short-circuit of
transformer is given in Cl. 4.1 of IEC 60076–5 (2000)
The symmetrical short-circuit current I = U/{÷3 ¥ (Zt+ Zs)} in kA
Where Zs = U2s/S in ohms
Us = rated voltage of the system in kV
S = short-circuit apparent power of the system in MVA
U = rated voltage Ur of the winding under consideration in kV
Solved Examples 607
Zt = (zt ¥ U2r)/(100 ¥ Sr) in ohms

zt = short-circuit impedance of transformer in percentage
Sr = rated power of transformer in MVA
The average temperature q1 attained by winding after short-circuit
shall be calculated by the formula q1 = q0 + {2 ¥ (q0 + 235)}/[{106000/(J2
¥ t)} – 1] for Cu winding.
where q0 = initial winding temperature in °C
J = short-circuit current density in Amp/mm2
t = short-circuit duration in seconds
HV winding
Zs = 212/9500 = 0.046 ohms
Zt = (12 ¥ 212)/(100 ¥ 50) = 1.06 ohms
I = 21/{÷3 ¥ (0.046 + 1.06)} = 10.96 kA line current
= 6.3 kA phase current
J = 6.3 ¥ 1000/350 = 18 Amp/mm2
q0 = 50° C + 55°C = 105°C
q1 = 105 + {2 ¥ (105 + 235)}/[{106000/(182 ¥ 3)} – 1]
= 111.3°C (Allowed 250°C)
LV winding
Zs = 11.52/500 = 0.2645 ohms
Zt = (12 ¥ 11.52)/(100 ¥ 50) = 0.3174 ohms
I = 11.5/{÷3 ¥ (0.2645 + 0.3174)} = 11.4 kA line current
J = 11.4 ¥ 1000/1050 = 10.9 Amp/mm2
q0 = 50°C + 55°C = 105°C
q1 = 105 + {2 ¥ (105 + 235)}/[{106000/(10.92 ¥ 3)} – 1]
= 107.3°C (Allowed 250°C)
C HAPTER 29
Index
Acrylonitrile 67 Coil
Ambient, weighted 382, 384, 387, assembly 169
389 treatment 68
Amorphous steel 84 Cold rolled steel 86, 87
Ampere turn Combined fault scheme 463
lancing 205 Commissioning of transformer 447
diagram 207, 208 Compressibility 57
Analysis of gaseous incipient fault Condenser
450 bushing 563
Analytical method 159 grading 564
Angle ring 56 Conservator diaphragm type 264
Arc suppression reactor 483 Construction
Askarel, see Oil features of traction transformers
Auto transformer 23 489
Auxiliary limb 88, 91 of tap changers 148
Axial flux 177, 156, 160 Continuous transposed conductor
59, 62
Bell-shaped tank 272 Control of onload tap chagers 151
Breather 260, 265 Controlled proof stress copper 84
Buchholz relay 251, 253 Conventional drying 278
Bushing Converter
D.C. 530 transformer connections 531–
and cable sealing box 257 533
insulating material 565 transformer test 536–538
Cooling 177
Cable sealing box 257 classification 177
Capacitance 118 Copper silver alloy 84
distribution 118 Copper weight 210
CBIP 371 Core
CEA 370 fault 463
Chopped wave impulse 117 insulation 292
Clamp plate 92 less shunt reactor 476
610 Index
saturation 87 Epstein square 100

steps 199 Equipotentials 52
types 200 Equivalent circuit 16–20
Correction factor 414–416 Explosion vent 257
Cross grain direction 76
Crystallite 68 Failure electric circuit 166
Faraday’s law 6
Design features of traction Fault
transformers 489 core 463
Despatch 439 delection 267
Dielectric earth 459
dissipation interturn 461
loss 129 short circuit 456
test 297 winding 461
Differential protection 462 Feeder reactor 466
Diffusion coefficient 278 Filtration 280
Dimensions of coil 204 Finite element method 236
Directed oil flow 125, 126, 127, Finite difference method 131
178, 179 Flange joints 272
Dirichlet boundaries 133 Flash point 45
Dissolved gas 280 Flash test core insulation 292
analysis 450 Flashover 307
Driving force 195 Flexible copper conductor 66
Drying Forces
by stream line filter 447 electromgnetic 154–161
of transformer 446 mechanical 218
Duplex reactor 467 Frictional head loss 192, 194
Full wave impulse 117
Earth fault 459
Earthing transformer 483 Gap volume in reactor 481
Eddy current loss 21 Gas sealing conservator 260
Efficiency 22 Generator line reactor 466
Electric field poltting methods 130, Grain oriented steel 68, 87
131 Grapped core shunt reactor 477
Electrolytic grade copper 62
Electromagnetic Handling 440
forces in winding 154, 158 Harmonics of no-load current 304
Electromotive force 7 Heat carrier 281
Electrostatic Heat transfer in winding 125
field plot 134 Hi-B steel 69
shield 120 High tensile strength steel 81
Energy dissipation in reactor 406 High voltage disturbances 457
Index 611
Hot spot 386 winding 113

temperature 92 Interphase transformer 513
Hygroscopic 274, 279 Interpretation of oscillogram 318
Hysteresis loss 21 Interturn failure 461
Interturn fault 461
Ideal transformer 7, 10 Iron
Image method 212 loss 210
Impedance 205 test 249
Impregnated cellulose 84 weight 210
Impulse
calculation 218 Ladder network 122
chopped wave 306 Lamination preparation 101
full wave 306 Lamination sliting 102
generation 309 Laplace equation 131
recording 315 Leakage flux 212
test 305 Leakage reactance 15
circuit 309 Lightning impulse test 306
testing 305–320 Lightning surge 116
voltage measurement 305 Line trap 470
voltage wave shape 306, 309 List of standards 372–376
wave front 309 Load curve 385
wave shape control 309 Load loss 22
Induced EMF 6 Loss measurement 345
Induced over voltage test 298
Inductance 118, 121 Magnetic
distribution 118 leakage 14
Initial voltage distribution 118 yoke shunt 214
Installation Magnetizability 68
of bushing 443 Magnetization D.C. 529
of transformer 441 Magnetizing
Insulation current 7
ageing 379 inrush current 463
ageing law 381 Magnetostriction 68, 76, 78, 81
deterioration 378 effect on noise 76
major 128 Maintenance 449
minor 127 of transformer 449
pressboard 51, 128 Major insulation 128
resistance test 297 Measurement
Interfacial tension 43 of impedance 295
Interleaved of insulation resistance 297
joint 100 of load loss 295
612 Index
of no-load current 296 Oscillograph recurrent surge 308

of no-load loss 296 Over voltage test 342–345
of winding resistance 294 Overload effect 406
Mechanical force 218 Oxidation stability 46
Metal limiter 67
Metallic glass 84 P.D.
Method of potential connection 147 direct current 527
Minor insulation 127 direct current measurement 529
Mitred joint 100 measurement 321
Moisture content 44 detection 329
Moitsinger relation 381 Parallel operation 27
Motor startomh reactor 468 Partial discharge 321
Multiwinding transformer 122 Partially interleaved winding 114
Peltier effect 265
Neaumann boundaries 133 Permeability air 48
Need for standardization 370 Phase sequence 31
Neutral Pitch of transposition 63
couplers 483 Polarity 30, 31, 292
earthing reactor 482 test 292, 537
Neutralization value 45 Polymerization degree 47
Nitrile butadiene 66 Poly-vinyl acetal based enamel 62
Nitrile rubber bonded cork 66 Porcelain bushing 562
Noise 76, 304 Pour point 45
Vibration in reactor 479 Power frequency test 297, 298
No-load Preliminary test 292
current 10 Pressure relief valve 255
loss 21 Protection 456
Non-magnetic inserts 217 differential 462
Numerical method 159 earth fault 459
internal fault 462
Off circuit tap switch 138, 139 winding fault 461
Oil 34, 42, 47 Proximity effect 510
base stock 42 PTFE 66
filling 445
level indicator 256 Quality factor 30
maintenance 279, 445
preservation 258 Rabin’s method 212
On load tapchanger 141 Radial
Operating condition, abnormal 404 flux 212
Operation control of tapchanger force 155, 159
151 Ranga-Kutta method 120
Index 613
Ratio test 292 Solvent 281

Reactance 205–210 Specification transformer 368
Reactor Specification convertor transformer
for capacitor bank 469 535, 536
magnetically shielded 475 Stacking factor 87, 92
Reconditioning of bushing 450 Standard specification of trans-
Rectifier circuit 499 former 368
Recurrent surge oscillography 308 Stiffeners 230
Reed type switch 253 Stiffeners matrix 240
Regulating transformer 503 Stored energy 121
Regulation 21 Strain energy 237
Reinforcement of insulation 120 Stray loss 22, 211, 213, 216
Resistivity 45, 51 Stray loss control 214
Restricted e/f protection 461 Stress
Ripple 497 in pipes 192
Rotary transposition 112 relief of lamination 102
Roth method 163, 212 sensitivity 78
RSO technique 308 Structural steel 81
Super conducting material 84
Sandwich coil 208 Surge
Sealing 66 divertor 457
Separate source withstand test voltage behaviour 116
338 voltage in winding 116
Series reactor 463 Switching impulse 306, 307, 313
Series reactor general character Synchronizing reactor 468
471
Series reactor oil immersed Tank
coreless 472 stress 248
SF6 insulated 84 tests 244
Shielded layer winding 115 type 227
Short circuit 78, 456 Tap changer type 202
fault 456 Tap changing circuits 143, 144
forces 154, 168, 456 Temperature
inter-turn 461 alarm 253
temperature rise test 300 gradient 219
Shunt reactor 473 indicator 253
Shunt reactor general character rise limits 300
475 variation 384, 395
Skin effect 509 Terminal gear 136
Sludge and sediment 46 assembly 271
Smoothing reactor 469 Tertiary winding 25, 26
614 Index
Tests Types of traction transformers 486

impulse voltage 345
induced overvoltage 298 Utilization factor 93
core insulation 292
Vacuum
impedance 295
dehydration 280
insulation resistance 297
impregnation 280
load loss 295
Vector group 32
no-load current 296
Veneer 58
no-load loss 296
Vibration measurement 339
phase rotation
Viscosity 44
polarity 292
Voltage regulation, see Regulation
ratio 294
separate source voltage 297 Water cooled transformer 178, 182
winding resistance 294 Weighted ambient 384, 389, 392
of reactor 335 calculation 386, 397
Thermal Winding
conductivity 125 assembly 269
head 180 continuous disc 110, 111, 268
image 254 crossover 113
Thermosyphon filter 263 direction of coil 110
Third harmonic measurement fault 461
340 gradient 302
Three winding transformer 25 helical 110, 111, 268
Tie line reactor 467 insulation 204
Top oil rise limit 300 insulation resistance test 297
Traction transformer 493 polarity 292
Transductors 517 preparation 268
Transformation resistance test 294
ratio 9, 23 spiral 110
Transformer temperature rise 301
auxiliaries 251 treatment 279
life 378 types 201
utility factor 506
Transient over voltage 116 Young’s modulus 174
Transposed conductor 62
Transposition 112, 217 Zero phase sequence impedance
Tuning, filter reactor 482 303

Bhel - Transformers-Tata McGraw-Hill Education (2003)

Uploaded by

Copyright:

Available Formats

You might also like

Bhel - Transformers-Tata McGraw-Hill Education (2003)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bhel - Transformers-Tata McGraw-Hill Education (2003)

Uploaded by

Copyright:

Available Formats

TRANSFORMERS

Bharat Heavy Electricals Limited

Tata McGraw-Hill Publishing Company Limited

© 2003, 1987, Bharat Heavy Electricals Limited, Piplani, Bhopal, MP

No part of this publication can be reproduced in any form or by any means

This edition can be exported from India only by the publishers,

Published by Tata McGraw-Hill Publishing Company Limited,

The McGraw-Hill Companies

It is indeed a pleasure to see the revised edition of this book on

Dr. Sankar Sen

I am pleased to know that the book on Power Transformers, cover-

Dr. P.C.S. Krishnayya

I am happy to introduce the revised edition of the book on

Preface to the Second Edition

editorial committee headed by Shri T. K. Ganguli, an eminent

Preface to the First Edition

The book covers all the aspects of power transformers—basic funda-

3. Materials Used in Transformers 34

7. Electromagnetic Forces in Power Transformers 154

9.16 Stray Losses in Transformer 211

13. Drying and Impregnation 275

16.8 Alternative Approach for the Calculation

21. Rectifier Transformers 496

24.4 Quality Aspects 560

Designing an insulation system for application in higher voltage

teed parameters like percentage impedance, no load loss, load loss,

according to duty application, viz. current limiting, neutral

The need for a reliable and stable system is being increasingly

2.1 Working Principle of a Transformer

A transformer is a static piece of apparatus used for transferring

Figure 2.1 Diagram showing magnetic circuit and windings of

2.2 Elementary Theory of an Ideal Transformer

An ideal transformer is one which has no losses, i.e. its windings

Figure 2.2 Vectorial representation of induced emf of an ideal

fore, it produces self-induced emf in the primary. This self-induced

2.3 EMF Equation of a Transformer

Let N1 = Number of turns in primary

The instantaneous value of counter electromotive force e1 is

2.4 Voltage Transformation Ratio (K)

From Eqs. (2.2) and (2.3), we get

2.5 Ideal Transformer

We will consider two cases,

(a) Transformer on No-load

I0 = (Im2 + Iw2 (2.8)

Figure 2.3 No-load current of an ideal transformer (having core

copper-loss is negligibly small which means that no-load primary

(b) Transformer on-load

Figure 2.4 Transformer on-load.

passing through the core is approximately the same as at no-load.

Figure 2.5 Vector diagram of an ideal transformer.

If we neglect I0 as compared to I¢2 as shown in Fig. 2.5(c), then f1 = f2

2.6 Transformer Having Winding Resistance

An ideal transformer was supposed to possess no resistance but an

the two windings. The result is that:

(a) (b) (c)

Figure 2.6 Vector diagram of an ideal transformer (having core

2.7 Magnetic Leakage

Leakage flux Main flux

Figure 2.7 Leakage flux.